Lucas Fluegel and Nick Corvino team up to tackle Chinese biotech. Lucas is a visiting scholar at the Carnegie Endowment for International Peace, where he explores biotech and biosecurity policy. He did his Ph.D. research in biochemistry and bacterial genomics at the Scripps Research Institute.

China wants to be the world’s biotech superpower. But to understand how it got here, it’s best to start with its crown jewel: the WuXi companies.

The WuXi companies are the dominant biotech services consortium in China and have become the lightning rod of U.S. political wrath, most notably as an early target of the BIOSECURE Act.

When we say “WuXi,” we don’t just mean WuXi AppTec. Although this family of companies is often spoken about as if it were a single company, in reality, it is a group of companies comprised of WuXi AppTec (药明康德), WuXi Biologics (药明生物), and a set of tightly integrated businesses, all more or less under the same leadership but dispersed throughout the industry. Together, they are stronger than the sum of their parts, and form what we envision as the Empire of WuXi (hereafter just “Wuxi”).

The TSMC analogy is tempting, since just as TSMC manufactures chips for companies like NVIDIA and AMD, WuXi, instead of discovering and commercializing its own blockbuster drugs, it provides the services (chemistry, testing, manufacturing) that allow others to do so. And both have the ability to gut-punch the global economy if their employees stop coming to work.

But AI analogies, tempting as they are, can do more harm than good. TSMC sits at a true chokepoint, with essentially no major rivals. If you want cutting-edge chips, you go through Taiwan. But WuXi does not monopolize a single irreplaceable step in the biotech supply chain. In fact, it has strong competitors both in China and globally.

WuXi AppTec and WuXi Biologics are the third- and fifth- largest contract development and manufacturing organizations (CDMOs) in the world by revenue. The remainder of the top ten are all based in U.S. partner nations, including the top two of Lonza (a Swiss company) and Catalent (a U.S. company). So, if there are plenty of alternative companies in U.S.-aligned nations, why is WuXi such a bogeyman for the U.S.?

In the same way that China’s rare earth stranglehold matters because of where those minerals sit in critical supply chains, WuXi, with its unique corporate structure, is embedded at many layers of the biostack. It has accumulated a structural indispensability that is harder to replace than a single dominant manufacturer would be.

A 2024 survey by the Biotechnology Innovation Organization estimates that 79% of US biopharma companies have at least one contract with a Chinese CDMO or CMO. WuXi AppTec alone is estimated to be involved in roughly a quarter of all drugs used in the United States (according to WuXi). And an estimated 65% of WuXi AppTec’s total revenue comes from U.S.-based clients.

Even if the U.S. and its allies lead in certain sectors of biotech, the growing recognition that WuXi has embedded itself throughout the supply chain has raised concern about systemic dependency and the leverage that comes with it.

The U.S doesn’t have an easy way to address this. China’s specific advantages in biotech look less like control over a single node and more like what it achieved with its manufacturing sector. It is about process expertise, cost efficiency, labor and talent, and deep integration into global supply chains — perhaps more like BYD’s success in the EV sector. These are not easily reducible to export-controllable chokepoints.

The biotech landscape is much more diffuse than AI. And yet, perhaps because of the TSMC analogies, Washington has increasingly tried to map its AI playbook onto biotech, with early versions of the BIOSECURE Act explicitly targeting WuXi as it would a company like Huawei.

We’ll get to the geopolitics at the end, but let’s first explore where WuXi came from and why they are so unique, before returning to how U.S. policymakers might approach the emergence of Chinese biotechs.

The Origins of WuXi: A Chinese–American(?) Story

The seeds of the WuXi empire were planted at a moment when it was relatively easy to build companies that straddled the U.S. and China.

Wuxi’s founder, Li Ge (李革), is emblematic of a particular early-2000s generation. Educated at Peking University, he earned his Ph.D. in organic chemistry from Columbia University and went on to become a founding scientist at Pharmacopeia, a U.S. biotech built around combinatorial chemistry.1 By the late 1990s, Li was fully embedded in the American biotech world, becoming a naturalized U.S. citizen. He is also very charismatic and speaks fluent English, meaning you’d often see him on TV segments talking to Western reporters:

And yet, like many in that cohort of returnees — the so-called “sea turtles” (海归) — he felt pulled back to China.

Around 2000, during business trips back to China, Li noticed something that is fairly obvious in retrospect but was underexploited at the time. China had a large pool of well-trained, low-cost chemists, while Western pharma companies were steadily increasing their appetite for outsourced R&D, driven by the rising cost and complexity of drug development. Bringing a new drug to market was getting more expensive as the low-hanging fruit had already been picked, older blockbuster drugs were losing patent protection, and the revenue that funded new research was starting to dry up. Outsourcing was the pressure valve, letting companies chase more drug candidates without expanding their own overhead. At the same time, China’s entry into the WTO and improving IP protections were making it newly viable to plug into the global pharmaceutical system. As Li later put it:

“Around 2000, as China prepared to join the World Trade Organization, intellectual property protection in the Chinese pharmaceutical industry significantly improved. I realized that Chinese pharmaceutical companies definitely needed to develop new drugs.”

He founded WuXi PharmaTech in 2000 with his wife, Zhao Ning (赵宁). Pharmacopeia, his former U.S. employer, became its first client.

From the beginning, WuXi PharmaTech was built as a cross-border company. It served Western customers, adopted international standards, and quickly oriented itself toward global markets. In 2007, it listed on the New York Stock Exchange, becoming one of the first Chinese biopharmaceutical companies to do so. However, WuXi PharmaTech later restructured, delisting from the NYSE in 2015 before relisting WuXi AppTec on the Shanghai Stock Exchange and Hong Kong Stock Exchange in 2018, alongside a separate Hong Kong listing for WuXi Biologics (药明生物) in 2017 and, more recently, WuXi XDC (药明合联).2 Wuxi made a series of correct bets on when to embrace the Chinese and American markets, respectively. Even today, although transparent data post-BIOSECURE is scarce, an estimated two-thirds of WuXi’s revenue comes from U.S.-based customers.

A key inflection point for WuXi is the 2015 reform of China’s drug review and approval system. By decoupling drug approval from manufacturing and encouraging outsourced production, the reforms accelerated a feedback loop: more innovative drugs → more R&D → more outsourcing → more innovative drugs, and so on. WuXi expanded aggressively to meet that demand and become the titan it is today, including earlier moves like its 2008 acquisition of a U.S.-based AppTec business, which gave it both new capabilities and a physical foothold in the American market (and the name of its most famous company, WuXi AppTec).

WuXi was not alone in embodying this Chinese-American model. Asymchem (凯莱英) was founded by a Western-trained Chinese scientist who returned to Tianjin and built a contract services platform. Porton (博腾), based in Chongqing, likewise evolved into an internationally oriented pharma services company with a large U.S. footprint.

For years, this dual positioning was an asset. The intertwinement of the U.S. and Chinese biotech systems was not accidental but foundational to WuXi’s rise. Western pharma outsourced to China for cost and scale; Chinese firms like WuXi grew by serving those needs. You could argue that this was exactly the outcome the U.S. wanted before it realized how powerful China would become.

That equilibrium has since come under strain. In early 2024, after being named in the initial BIOSECURE Act proposals, WuXi’s stock plunged sharply, wiping out tens of billions in market value in a matter of days. Although some of those losses have since been partially recovered, WuXi is now a target of the U.S., and its future is highly precarious.

A generation of Chinese biotech companies emerged from this earlier era of integration, commercializing Western training and global demand through China’s industrial base. But WuXi remains the most internationally salient and successful of them all.

Why?

What Makes WuXi So Good?

Li’s vision for WuXi’s role in the pharma business ecosystem was explicit from early on. WuXi was not meant to be a traditional drug company, but an enabling platform for global innovators. Rather than designing drugs, they would build the infrastructure needed to quickly find and develop them. The novelty of this business model was not simply exploiting wage arbitrage — U.S. and European pharmaceutical companies already knew how to outsource chemistry. Instead, Li’s key insight was to reframe the role of contract R&D in the drug development process.

Traditionally, outsourcing drug companies would partner with different contractors for each step of drug development. WuXi provided an enticing alternative. Instead of contracting one company to test the initial drug, another to optimize its potency, and another to manufacture it at commercial scale, drug companies could work with WuXi through the entire pipeline.

Li would later define this approach as an “open-access platform” (开放式平台). Unlike more siloed competitors, WuXi was committed to “following the molecule” as it progressed from the research laboratory to regulatory approval and commercialization. This business model would later be codified as a “contract research, development, and manufacturing organization” (CRDMO) and copied by other companies.

This approach is a win-win for both parties. For the drug developer, it minimizes the need to switch between different corporate ecosystems, eliminating the inefficiency of juggling multiple contracts and ensuring each partner is up-to-speed. For WuXi, it incentivizes customers to stay “stuck” to their services for years, leading to predictable business and access to the revenue scaling that occurs as the drug progresses towards commercialization. Given the immense uncertainty involved in pharmaceutical development, this level of stability for provider and customer is extremely attractive.

WuXi doubles down on this model by targeting a “long tail” of biotech customers. Rather than limiting themselves to massive deals with the pharmaceutical giants, they target many small- and medium-sized firms. With more limited resources, these small companies benefit particularly from the cost efficiency of WuXi’s end-to-end services, which then locks them into the pipeline. Their sheer number and diversity also diffuse the risk of major damage from any one customer pulling out. Furthermore, research by consultancy firms has shown that these smaller companies tend to produce more innovative drug leads than their big pharma counterparts. WuXi is therefore able to link itself to these disruptive — and therefore lucrative — products early on. These strategic decisions have given WuXi a “strong, diverse, and sticky customer base.”

Does WuXi have a technical moat?

Importantly, however, these technologies didn’t originate from WuXi labs. So, unlike the TSMC analogy, there is not a WuXi-specific technological moat around their services. Instead, WuXi’s biggest competitive advantage lies in their integration across the technology stack.

Indeed, a quick scan of their advertised capabilities reads like a catalog of the hottest frontier capabilities in drug development. A company can use WuXi’s DNA-encoded libraries to quickly scan for usefully potent molecules, including with options to avoid sharing IP. Biomanufacturing for complex biologics has been standardized and optimized, with new methods being deployed to further boost productivity at scale. In-house expertise in finicky drug types like peptides (including GLP-1s), antibody-drug conjugates (an expanding class of mainly anticancer drugs), and monoclonal antibodies (of COVID-19 treatment fame) expands the customer base they can serve. And, of course, AI and automation are being deployed throughout the pipeline.

Most biotech and pharmaceutical firms lack the resources and expertise to deploy these advanced biotechnologies in-house. But WuXi’s comprehensive and integrated platform offers them the access and support needed to compete at the technological frontier. A positive feedback loop is born as WuXi aggressively invests in further optimization and expansion, and the platform becomes even more attractive to the next wave of ambitious firms.

An excellent example of WuXi’s ability to adopt and deploy new technologies is their development of the “scale out” paradigm for manufacturing biologic drugs.

Biomanufacturing – the use of a living system or its parts to produce a good – is central to many of WuXi’s higher-end pharmaceutical manufacturing processes. The key step in a biomanufacturing process is growing the organism that makes your desired product in a large vessel, called a bioreactor. Traditionally, scale-up of these processes would proceed linearly, moving gradually to larger bioreactors until the necessary commercial scale is attained.

But the conceptual simplicity of this approach hides many downsides. Bigger bioreactors change the physical processes within, often leading to unexpected engineering problems like poor stirring or slow oxygen transfer. Product yields are compromised, requiring expensive and time-consuming optimization at each stage. Simultaneously, the capital expenditure and financial impact of contaminated batches scales with the bioreactors.

WuXi sidestepped these challenges. In place of building >20,000-liter tanks, they run multiple 2,000- to 4,000-liter reactors in parallel: scaling out instead of up. By doing so, the same proven operational conditions are used at small and large scales. Making more or less of a product requires no additional engineering — simply add or subtract bioreactors. The separation of one production run into several batches also ensures that one contamination event does not spoil the entire campaign. WuXi’s adoption of single-use disposable systems that don’t require meticulous cleaning between runs has simplified operations even further. Though not the first to develop these technologies, WuXi was the first to pioneer it as the backbone of a commercial-scale manufacturing capacity.3

WuXi’s China Advantage

Deploying these suites of frontier technologies and large-scale manufacturing facilities is expensive: WuXi Biologics’s massive Singapore facility reached a price tag of $1.4 billion. But some of this financial pain is offset for WuXi by the favorable political and economic landscape of China’s science and technology sector.

The most critical advantage is the Chinese workforce. Chinese universities produce dramatically more STEM Ph.D. graduates than their U.S. counterparts. WuXi capitalizes on this geographic concentration with targeted training programs that attract top candidates and develop company-specific skills. WuXi also invests in training workers at every level of the production process, including the technicians and operators running factory floors. This is precisely the kind of vocational and technical workforce development that the U.S. has chronically underfunded and undervalued. Because this highly skilled Chinese talent is often half the cost or less than Western equivalents, companies like WuXi can deploy larger teams to shorten timelines and overcome obstacles.

WuXi also benefits from China’s established excellence in advanced manufacturing. Because China largely controls global production of raw materials and active ingredients for small-molecule pharmaceuticals and is rapidly domesticating the supply chain for biologics, domestic companies benefit from easier sourcing and more resilient supply chains. This colocalization directly translates into accelerated procurement and lower overhead costs.

These advantages are compounded by the central government’s aggressive championing of biomanufacturing, such as labeling biomanufacturing a national priority and doling out subsidies.

Overall, this investigation shows that WuXi’s success is not a result of some unassailable technological lead in a core competency area. Instead, the well-rounded profile of the company means there is no singular source of advantage. This fact presents an unusual problem to concerned policymakers, who have been struggling to figure out how to deal with WuXi for years.

The Geopolitics of WuXi

Led most prominently by the National Security Commission on Emerging Biotechnology, the U.S. is racing to determine how to maintain its competitive edge in biotech in the face of rising Chinese pressure. The threat of losing the advantage in innovation or a cutoff of basic medicines has policymakers searching for options. The size and success of WuXi has naturally caught their attention.

The BIOSECURE Act is the most notable move. It prevents the use of federal dollars to pay for goods or services from biotechnology companies of concern. In the earliest versions of BIOSECURE, WuXi AppTec and WuXi Biologics were both explicitly targeted. By pushing U.S. companies away from contracting with WuXi, it was hoped that new and more U.S.-aligned firms would step up to fill the gap.

However, the explicit naming of companies was abandoned in the final version of the Act that was passed as a part of the 2026 NDAA. Given how much pain this would have cost U.S. firms, since WuXi is embedded in a quarter of all drugs in the U.S., quitting cold turkey would have been painful.

Unlike with restricting AI components, where slowing progress would be felt years after implementation (and might even be welcomed by Americans already anxious about the technology), the costs of disrupting access to cancer drugs or GLP-1s would be immediate and personal for Americans.

Instead, companies of concern are determined by a deliberative process led by OMB or inclusion on the DoW’s 1260H list of “Chinese military companies”. Unusually, the 2026 version of this list was released for only a short time before being quickly removed from the Federal Registrar. WuXi AppTec was included on this since-removed update, despite being absent from previous versions. So, though the pathway is different, it does seem that BIOSECURE is poised to target WuXi after all.

Implications for U.S. Policy

The U.S.’s policy response to WuXi is an interesting piece of the broader U.S.-China biotech puzzle. Here are a few loosely-held takes:

Take #1: It seems the U.S. policy apparatus is using this company-banning/targeting approach because of its familiar success from AI. But, because most of WuXi’s advantages don’t come from any particular technology lead, does this approach really apply in this situation?

Take #2: The U.S. is quite concerned that China is “catching up” in biotech despite spending far less on relevant R&D:

To us, this suggests that the geostrategic competitive pressures we want to address aren’t primarily about money. If Chinese firms are making important moves with only a fraction of our budget, then whatever advantages they’re exploiting are probably not going to wither away if we further restrict funding. Instead, it seems like we need to think more creatively about how we can race further ahead instead of only worrying about how to slow down our competitors.

Take #3: It’s unrealistic to expect the U.S. to unilaterally dominate every layer of the biotech stack. The U.S. remains a global powerhouse in biotech, occupying advantageous positions across the entire technology stack. But, China is a massive country with a well-educated workforce that has decided to focus major investments into biotech – it’s inevitable that they will become an influential player. Perhaps the right question isn’t how we eliminate Chinese participation in biotech globally but which specific capabilities, if ceded, we could live with.

No single country is waiting to absorb WuXi and China’s cheap and diffuse biotech role. India has a large base of FDA-approved facilities, competitive costs, salaries about half of China’s, and a large and growing Ph.D. pipeline. It has thus received a surge of inquiries from U.S. pharma eager to diversify away from China. However, most of India’s strength is concentrated in small molecule generics, a very different skill set from the complex biologics manufacturing that makes up so much of WuXi’s value. South Korea’s Samsung Biologics is strong on biologics (rivaling WuXi Biologics), but weaker on the small molecule CRO and chemistry services where WuXi AppTec has built its deepest moat. No single country or company can replace all of the different roles WuXi plays, but if the U.S. leveraged its multilateral relationships to build a coordinated alternative across trusted partners, that would be its best shot, something Trump 2.0 has moved against.

The uncomfortable truth is that a U.S. biotech industry fully decoupled from China would be a slower and more expensive one. Policymakers need to be honest with themselves about that tradeoff, unless they think Americans will be fine with fewer cancer drugs for the foreseeable future.

Yesterday, my colleague Aqib Zakaria published an estimate of China’s supply-side compute capacity. By tallying chip shipments, smuggling reports, domestic production, and estimated Western cloud access, he arrived at ~2.7 million H100-equivalent GPUs.

Today, I’ll try to approach the same question from the demand side. Rather than counting chips, I’m counting workloads to estimate how much compute China’s AI ecosystem needs. The demand-side approach is less precise than the supply-side (which means much more vibes-based guessing from me), but it offers a cross-check on whether the supply-side figure holds up.

The number I landed on is ~2.8 million H100e, which is nearly identical to Aqib’s estimate — though it’s entirely possible we’re both wrong in ways that happen to equalize. (We importantly did not share our numbers until after they were calculated!)

Why does this matter? For one, export controls on advanced chips are only as good as our understanding of what those chips actually enable. If policymakers are debating whether to tighten restrictions on H200s or close cloud compute loopholes, they should probably have a concrete sense of what China’s AI ecosystem actually demands. Looking at demand could further allow us to infer how much compute Chinese companies are renting from Western cloud service providers.

The rest of this piece walks through how I got my number. In a nutshell…

I estimate China’s AI infrastructure requires roughly 237,000 H100e running continuously to serve all inference workloads, such as chatbots, enterprise APIs, recommendation algorithms, video generation, surveillance, and more. Dividing by 55% — my central estimate for how intensively deployed inference chips are actually being used, based on various conversations and the common wisdom of the utilization discourse — gives a minimum inference installed base of ~431,000 H100e. Add a dedicated training cluster of ~128,000 H100e, used episodically for model training and research, and you get a minimum total installed base of around 558,000 H100e. Scale up to account for chips in reserve, in transit, between runs, or not yet fully online, and you reach 2.8 million at 20% whole-fleet utilization (one again based on conversations and the common range proposed by others). The training cluster is handled differently from inference: rather than dividing by the 55% inference utilization rate, it’s derived from total annual training compute divided by available chip-hours per year at a higher utilization rate (80%) due to the increased efficiency of usage during training.

I’ll walk through each step in more detail below. If you want the full methodology, the complete methodology is available here: China’s AI Compute Demand.

Working through this surfaced a few things I’d really like to know more about:

• GPU utilization rates. Choosing the correct utilization percentage is the most important factor in the calculation, and quite contested. Based on multiple conversations with others and the common wisdom in the literature, I use 40-70% for how hard deployed inference chips are actually working, and 10-30% for the share of China’s total installed fleet operating at a given moment (central estimate: 20%). But a one-percentage-point shift in that whole-fleet figure moves the final number by roughly 186,000 H100e, so being accurate is incredibly important, and more research would be useful here.

• Enterprise AI adoption. I estimate Chinese enterprise API usage at ~40% of US levels, which is the largest single inference category in my estimate. But neither Chinese nor Western AI companies disclose much about enterprise usage, which is frustrating.

• Surveillance and recommendation algorithms. How much compute do surveillance and recommendation algorithms require, and is the quality of compute comparable to H100e or more simple workload processing?

• Use cases I missed entirely. The limitation of demand-side analysis is that there’s always a category you didn’t think of.

Given the nature of this task, there are bound to be mistakes. But I hope this BOTEC serves as a framework for others to build on, poke holes in, and improve. If any of my numbers seem wildly off, reach out at nick@chinatalk.media or let me know in the comments.

Compute by Use Case

The chart below breaks down China’s total AI compute demand by workload category.

Inference

We’ll get to the biggest category — training — in a moment. But first, here are a few inference findings I found interesting.

Casual users barely register, even at an enormous scale. CNNIC put China’s generative AI user base at a whopping 602 million as of December 2025, though that figure counts people who touched an AI-powered feature even once, so it’s more of a ceiling than a headcount. I believe a more useful working estimate is around 175-240 million regular active users generating meaningful compute load, with perhaps 300 million more in the casual tier: people who send a quick translation, tap a Doubao suggestion, or use an AI-powered filter. At roughly 1,000 tokens per day for a handful of short queries, that entire casual population runs to about 1,000 H100e, less than a percent of total inference demand. A single enterprise customer burning through a trillion tokens a year contributes more compute than tens of millions of casual daily users combined.

Enterprise API is the largest single inference category. I estimate domestic enterprise API at roughly 50,000 H100e, or about 21% of total inference. My biggest anchor here is OpenAI’s 2025 Enterprise AI Report [pdf], which discloses around 200 organizations exceeding 1 trillion tokens per year and roughly 9,000 exceeding 10 billion. Treating those at their floor values and converting at roughly 0.8 million tokens per H100e-hour gives a US enterprise inference floor of about 41,000 H100e from OpenAI alone, and that’s before adding Microsoft (15 million paid seats), Anthropic, Google, and internal deployments (which I essentially estimate according to a vibe of how successful I think their enterprise AI endeavors are in respect to OpenAI). China’s enterprise AI ecosystem is growing — Alibaba Cloud’s Tongyi platform reportedly reached 300,000 enterprise customers in 2024 — but is earlier-stage than the US and has fewer users. Enterprise software in China tends to have shallower AI integration, fewer automated workflows, lower per-organization token volumes, and less of the deep API-level embedding into business processes that drives compute at US firms like the ones in OpenAI’s enterprise cohort. I’ve therefore estimated China at roughly 40% of total US enterprise demand, which gives the 50,000 H100e figure, but I could see the number being anywhere from 20K-70K.

Surveillance, military-adjacent AI, recommendation algorithms, and unknown unknowns are collectively about 32% of total compute, but the hardest to pin down. For instance, recommendation systems (Douyin, Taobao, Kuaishou, Pinduoduo) run continuous real-time ranking inference 24/7, with little overnight demand drop. I’ve put these at 30,000 H100e, though much of this workload runs on edge silicon and specialized accelerators rather than H100-class hardware. The miscellaneous category — government surveillance AI, military-adjacent applications, industrial systems — I’ve put at 45,000 H100e. This is the number I’m least confident in. Real-time video reidentification across 600-700 million cameras is computationally intensive (backend aggregation and reidentification could require datacenter-class hardware even when edge NPUs handle the initial detection), and military AI usage leaves essentially no public trace. Of all the variables I could wildly over/under-predict, it is most likely one of these.

There are a couple of other variables I find interesting that you can see in the full breakdown, such as international users on Chinese servers (particularly DeepSeek’s large international user-base, which runs inference on Chinese hardware) and domestic and international video generation, which has grown into a meaningful compute category as platforms like Kling and Seedance scale up.

Training

Getting training compute onto the same footing as inference requires a different methodology. Inference is naturally expressed as chips running right now. Training is episodic; a cluster runs flat-out for weeks or months, then sits idle or is redirected towards inference between runs.

So instead of asking, “How many chips are active right now,” I asked, “How many GPU-hours does China consume on training per year, and how many dedicated chips does that imply?” My anchor is DeepSeek V3’s disclosed training run of 2.788 million H800-GPU-hours, adjusted upward by 1.5x for likely underreporting (I believe Chinese labs have strong incentives to minimize disclosed compute, since it reinforces the efficiency narrative and understates chip needs under export control scrutiny). From there, I built up tiers: five major labs (ByteDance, Alibaba, Baidu, Huawei, Tencent), each running multiple frontier LLM runs per year with overhead for ablations, RLHF, and failed experiments; ten mid-tier labs; five state and military-adjacent actors; plus video model training, image models, and AV simulation. Total annual training budget: roughly 298 million H100e-hours.

To convert that into a dedicated cluster size, I assumed training clusters are active for about four months per year at 80% utilization while running, yielding around 2,336 available hours per chip per year. Dividing 298 million by 2,336 gives roughly 128,000 H100e as the implied dedicated training pool. I also added ~50 million H100e-hours for research and experimentation beyond production training runs — architecture tests, scaling-law experiments, RL paradigm research — which pushes the total up from what a pure production-run accounting would suggest.

But not all compute is created equal. China’s training clusters require their best chips, like the H800s and smuggled Blackwells. This means China may have ample compute for inference at scale while still facing a meaningful constraint on how fast it can train the next generation of frontier models.

The Final Number

Now that we have inference and training accounted for, we need to bridge from “compute actively running” to “chips China actually has installed” — and that requires two separate utilization adjustments.

The first is installed base utilization: the share of deployed inference chips that are processing requests at any given moment. This isn’t 100% even for a busy cluster — there are overnight troughs when query volume drops, headroom kept for traffic spikes, downtime when logic is waiting for memory, and chips idling between requests. I use 55% as my central estimate, which sits at the midpoint of what I think is the plausible 40-70% range based on an aggregate of the existing literature and conversations with experts. At 55%, the 237,000 H100e running continuously implies an inference installed base of around 431,000 H100e. Add the training cluster of 128,000 H100e — which, as noted above, has utilization already baked into its calculation — and you get a minimum total installed base of roughly 558,000 H100e. This is the floor: the fewest chips China could have while still producing the workloads we observe.

The second is whole-fleet utilization: the share of every chip China has ever purchased that is doing anything at a given moment. This is structurally much lower than inference-serving utilization because it includes chips that are newly procured and still in transit, recently installed clusters still ramping up, training hardware sitting idle between runs, and reserve capacity held for future demand. China has been procuring AI hardware at an extraordinary pace, and a significant fraction of that hardware is, at any given snapshot, somewhere in the pipeline rather than serving workloads. I use 10-30% as the plausible range, with 20% as the central estimate.

Dividing our median estimate for the minimum installed base of 558,000 H100e by 20% gives approximately 2.8 million H100e as the implied total Chinese AI chip stock. At the pessimistic end of the utilization range (10%), that number rises to 5.6 million; at the optimistic end (30%), it falls to 1.9 million. The sensitivity is dire, and it’s why utilization assumptions deserve more dedicated research.

2026 growth scenarios

Global AI token consumption has grown at roughly 5× annually, and there’s no particular reason to think China is an exception. To get a sense of where things could be by the end of 2026, I’ve modeled three scenarios: 2×, 5×, and 10× growth applied to the inference pool, with the training cluster held flat as I am unsure if training will become more or less compute intensive in the future.

At 2× the implied whole-fleet installed base roughly doubles to ~5 million H100e; at 5× it reaches ~11 million; at 10× it climbs to ~22 million. For context, Epoch AI estimates current global high-end AI compute capacity at 20 million H100e, meaning the upper end of these scenarios implies China’s AI fleet alone approaching the scale of today’s entire global stock. That’s probably not happening by the end of 2026, but it illustrates why demand-side analysis matters: the numbers we’re dealing with today are large, but the trajectory is what’s really consequential for thinking about export controls, chip procurement, and the balance of AI capacity between the US and China.

Thinking about the future is ultimately what I hope this BOTEC is useful for. A supply-side count tells you what China has. A demand-side count tells you what China needs today, and what it will need in the coming years.

You can read the full methodology here: China’s AI Compute Demand

If you want to publicly launch an AI product in China, you need to get on the government’s safety list. You can see all 6,000+ approved companies in plain sight (courtesy of Trivium: excel file). But what’s less clear is how to actually get registered. Some vendors claim they’ll do it for you. Others claim they can do much more.

Let’s take a closer look at the emerging marketplace for AI safety services in China.

We’ll begin with the cottage industry of online vendors promising to help companies navigate the filing process, then turn to the more formal third-party safety firms positioning safety as a full-fledged business model. Finally, we’ll examine how the West tends to frame safety and technological progress as opposing forces, a tension far less pronounced in China, before turning to what the rise of agentic AI could mean for the scale of China’s safety industry.

Subscribe now

For Context

Any company deploying generative AI services with “public opinion attributes or social mobilization capabilities” has to file with the Cyberspace Administration of China (CAC). Before you scale to the public, regulators need to be satisfied that your model won’t say the wrong things or “violate core socialist values.” You have to make sure your model won’t claim Taiwan is an independent country or explain what happened in 1989 in Tiananmen Square.

There’s been good coverage on the registration process from the Oxford China Policy Lab, Wired, and Trivium.

The CAC publishes its list of registered companies periodically, with approved models from giants like Baidu and ByteDance to startups you’ve never heard of. What the CAC doesn’t clearly explain is how to actually get on that list.

There have now been more than 6,000 filings. Alibaba alone has 67 products and algorithms registered. Inspur, a company I had admittedly barely heard of before this, shows up with 28 registrations. (Inspur is China’s largest server manufacturer and the world’s third-largest, specializing in AI servers and GPU systems that train large models!) Huawei has ten. DeepSeek has three.

The core requirements seem demanding: a 100+ page Safety Assessment Report detailing training data sources and security measures; a keyword interception list of at least 10,000 blocked terms covering 31 risk categories (political sensitivity, violence, discrimination, etc.), and the ability to appropriately answer a gauntlet of sensitive questions. According to the WSJ, this includes running a database of 20,000 to 70,000 questions testing whether the model answers appropriately, refuses the right questions, but also doesn’t over-refuse normal queries.

For large, well-resourced companies, I doubt the CAC requirements are a major inconvenience. But this got me thinking: a student startup building its first AI product faces the same regulatory requirements as Alibaba. The CAC doesn’t distinguish between billion-dollar incumbents and five-person founding teams; all these different players have to navigate the same compliance maze.

So how would a small team without regulatory expertise or deep pockets actually get through this?

Share

The Taobao Method

Search “AI evaluation testing” (AI评估测试) on Taobao 淘宝, China’s largest e-commerce website, and you’ll find a cottage industry of vendors advertising services that map onto CAC requirements. They rarely mention the CAC explicitly, but the implication is clear enough.

I messaged several of these sellers, posing as a Peking University student startup that had fine-tuned Alibaba’s Qwen model.

The first attempts didn’t go well. They immediately asked detailed technical questions about our product that I couldn’t answer, causing them to get skittish and cut off communication. But after Claude helped me concoct a more coherent story, I was able to move the conversations to WeChat, where we could discuss CAC filing more directly.

Prices varied and seemed negotiable. Filing for recommendation algorithms or content moderation systems was notably cheaper than for full AI-generated content (AIGC). For AIGC — the comprehensive safety assessment required for generative AI services with “public opinion attributes or social mobilization capabilities” — quotes ranged from ¥15,000 to ¥80,000 (roughly US$2,000–$11,000).

What these companies offer is essentially full-service compliance. You handle the technical work; they handle the paperwork and regulatory interface. As one vendor explained:

“We are responsible for writing the materials for the large-scale model registration, while you are responsible for optimizing the model. The writing period is 15 working days, and we complete the large-scale model registration materials. The CAC will review them for approximately 4 months, depending on the local review timeline. If the materials are ultimately rejected, the CAC will not accept them, and the large-scale model registration will not be approved, and we offer a full refund.”

Alternatively, instead of a refund, some vendors offer to revise the materials until they satisfy the review requirements. The risk, however, is time. An AIGC filing typically takes two to five months for review. If the application ultimately fails, you may have to restart the process, stretching the timeline possibly to a year before you can officially launch. In AI terms, that kind of delay can feel like an eternity, with your hot product today facing the risk of becoming obsolete.

Subscribe now

When I asked whether we could just do it ourselves, I usually got this sort of response:

“The process is quite complex. You can try it yourself, but it will take time. In some industries, the required documentation can consist of over 500 pages.”

These vendors tried to imply that, unlike me, they had special information on how the CAC review process actually works: how to structure submissions, what common rejection points look like, and how to streamline back-and-forth with regulators.

After enough of these conversations, I realized that the financial and time costs of filing probably barely register for major AI labs. But imagining myself as part of a scrappy college startup, the process felt more daunting. Tens of thousands of yuan, not to mention months of review time, is not trivial when you are operating on a thin runway.

There is, however, the possibility of recouping some of that expense (Ch. 2, Section 3; p. 24 onwards). In many jurisdictions, successful CAC filing has become a prerequisite for accessing local government support programs. Cities and districts across China now tie model-registration status to one-time rewards, R&D reimbursements, compute subsidies, or model vouchers. In some cases, the headline figures run into the hundreds of thousands or even millions of RMB. For firms that qualify, the ¥15,000–¥80,000 spent on filing assistance can look less like a regulatory tax and more like a down payment on industrial policy eligibility.

But that support is far from automatic. Filing is usually a necessary condition, not a sufficient one. Many policies seem to apply only to first-time registrants, impose minimum parameter thresholds, cap annual payouts, require local incorporation, or distribute funds on a competitive, merit-based basis. Subsidies can meaningfully offset compliance costs, but they are neither guaranteed nor universally available. For smaller firms in particular, counting on government support to balance the books still looks like a gamble rather than a certainty.

Official Third-Party Safety Services

What if you’re a company that wants more than just a random online vendor to do your AI paperwork for you? A more formal layer of third-party firms has emerged in China to shepherd models through their safety/compliance journeys, and they do more than just help you pass the CAC requirements.

Firms like RealAI aren’t just trying to sell you on passing the CAC requirements, though they’ll do that for you if you ask. They also market end-to-end safety infrastructure: adversarial testing, robustness evaluation, content filtering, post-deployment monitoring, and broader controllable AI engineering. BotSmart (博特智能) bundles AIGC compliance with explicit “ideological alignment” testing and even deploys its own model to evaluate the outputs of other models.

Share

Baidu, ByteDance, and NetEase have all built out similar offerings, often by expanding the scope of pre-existing cybersecurity products. Zhipu AI has publicly stated that it uses NetEase’s services for pre-deployment dangerous capability assessments (Appendix C of this pdf), and SenseTime has signed a cooperation agreement with RealAI (Page 57 of this pdf).

For the best round-up of this space, see the final section of Concordia AI’s State of AI Safety in China (2025) year-end report, “Safety as a Service.”

These firms appear to be more ‘safety-pilled’ than simple compliance shops. RealAI, for instance, often publishes high-quality papers on AI safety. Their services extend to interpretability research, AI’s moral point of view, and loss-of-control scenarios, not simply passing government tests. There’s no telling how many customers they have (I asked, and they wouldn’t tell), and these companies also have many other business streams completely unrelated to AI safety. (BotSmart sells an AI pen, which is a writing device that uses built-in sensors and artificial intelligence to perform tasks like translating text or digitizing handwriting.) But these companies are hoping the safety market will grow as AI becomes more transformative and more of a headache for the Chinese government.

The Market Dynamics for Safety

What makes these third-party safety companies interesting is not just what they sell, but why the market exists at all.

In the West, governance vendors like Holistic AI or Credo AI help enterprises document risk and prepare for frameworks like the EU AI Act. Evaluation startups such as Haize Labs or Patronus AI specialize in red-teaming and scalable oversight. But these businesses are largely capitalizing on voluntary (or at least not mandatory) demand. They target companies worried about liability, reputation, possible future regulation, or those that simply believe in safety and are willing to spend on it absent any requirement to do so.

Much of the deeper safety work, meanwhile, is philanthropically funded, meaning it operates outside normal market logic. Safety doesn’t need to be profitable if it’s underwritten by foundation grants and EA-adjacent donors. The US government, meanwhile, has treated AI safety as something industry should sort out for itself, a posture Trump 2.0 has only reinforced. When the state doesn’t set the terms, the market does, and markets have little patience for those asking them to slow down.

This may explain why Western AI discourse has hardened into such a fierce binary, where caring about safety all too often reads as indifference to progress. In China, that dichotomy feels less pronounced, where both AI safety and market direction are assumed to be the state’s responsibility (though I’m sure there are internal battles between different government factions).

It would be a mistake, however, to read this exclusively as a more ambitious safety culture. Much of what is construed as safety in China is closer to compliance with ideological requirements than deeply mechanistic or ethical scrutiny, and thus, the safety discourse is also less fractious, partly because it sidesteps more fundamental safety questions.

Subscribe now

Product Market Fit

Compared to the West, China’s current ‘safety’ industry enjoys a much more concrete product-market fit. The CAC filing regime creates an immediate, regulator-facing bottleneck for publicly deployed generative AI. In effect, regulation precedes and shapes the market. Safety becomes not just best practice, but a prerequisite for launch, and could scale dramatically if regulators expand scrutiny toward more complex risks, such as agentic behavior, systemic misuse, and CBRN (Chemical, Biological, Radiological, Nuclear) risks.

BotSmart makes the pitch boldly, if not a bit ridiculously, in this white paper:

“According to industry data, the size of China’s AI safety market exceeded 89 billion yuan in 2024, is expected to surpass 113 billion yuan in 2025, and could reach 242 billion yuan by 2028, implying a compound annual growth rate of 22.3%. This growth is largely policy-driven. The Interim Measures for the Management of Generative Artificial Intelligence Services establish a principle of “mandatory review before launch,” turning AI security into a rigid, unavoidable requirement for companies.”

I’m skeptical of this prediction, not just because of its inflated numbers, but also its assumption of an inevitable increase in safety regulation.

For instance, China’s open-source culture means companies can build on existing Chinese models whose base weights have already cleared regulatory review, reducing the marginal compliance burden for additional companies. (This would be harder in the US, where leading models are proprietary and each firm would have to satisfy requirements independently.)

Furthermore, Chinese regulators have so far focused narrowly on political and social content control. CAC rules and enforcement rarely emphasize frontier concerns like CBRN misuse or misalignment risk, and weak performance from the top Chinese AI companies on such benchmarks hasn’t elicited much of a response. If that posture continues, demand for ’‘deeper’ safety services may remain limited.

That said, this framework may start to strain with the rise of AI agents.

Agents

Up to now, agents lack a dedicated national regulatory regime and are generally subject only to provincial-level review. But systems that act autonomously across payments, logistics, or communications are harder to govern with keyword lists and static banks of test questions. Models that can browse the web, call APIs, or interact with other software systems may introduce new ways of upending China’s existing controls.

How regulators adapt their existing toolkit to agentic AI is an open question — one ChinaTalk will explore soon! For now, my guess is that the CAC will do what it usually does: sharpen liability rules and push the technical problem onto companies, much as it did with platform content moderation. In practice this means regulators don’t need to specify every prohibited behavior in advance; they can simply punish firms when something they don’t like slips through.

If this is the path they take, an AI company facing genuine criminal liability for emergent agent behavior will need evaluators who can actually probe those systems adversarially, not just run a keyword battery. That’s where third-party firms like RealAI and BotSmart could scale up and become integral players in the AI market, since the incentive to produce real safety evidence, rather than just paperwork, might finally kick in.

Anthropic managed to massively piss off both the DoW and China in the same week.

For context: On February 23rd, Anthropic was summoned to the Pentagon by Secretary Hegseth, who demanded Claude’s safety guardrails be stripped for unrestricted military use. That same day, Anthropic published a blog post accusing three Chinese AI labs (DeepSeek, Moonshot/Kimi, and MiniMax) of industrial-scale distillation. A few days later, Trump called them a “RADICAL LEFT, WOKE COMPANY,” blacklisted them from all federal contracts. Hegseth then said he would designate them a national security supply chain risk, which was a label previously reserved for foreign adversaries like Huawei. The distillation accusations, meanwhile, landed in China as hypocritical politicking, compounding the bad blood from Anthropic’s September 2025 restrictions on Chinese-controlled companies.

Anthropic now occupies an unprecedented political position: regarded in Washington as too woke to be trusted, and in Beijing as the most hawkish AI company.

The Irony

The most palpable emotion on Chinese social media is irony. Given Anthropic’s track record with China — banning Chinese-controlled companies, labeling China an enemy state in internal documents, and pushing hardest in Washington for compute restrictions on Chinese firms — Chinese netizens were not exactly sympathetic when the blacklist dropped.

Subscribe now

Anthropic, which has done more than any other Western AI company to frame China as a threat, may now be deemed the same “supply chain risk” designation historically reserved for Chinese companies like Huawei. The mockery lands harder given that just weeks earlier, Anthropic was being called “AI Thanos” (“AI灭霸”) after its February product releases wiped out software stocks (IBM down 13%, CrowdStrike down 6.5%).

But there’s a second level of political irony. The US government, which built its entire AI export control regime around the premise that democracies develop AI differently from autocracies, spent this week threatening a company with criminal prosecution for refusing to enable domestic mass surveillance and fully autonomous weapons, the exact use cases Washington spent years warning China would pursue. From America’s AI Action Plan, the Trump Administration’s policy roadmap for AI released in July 2025:

“AI systems will play a profound role in how we educate our children, do our jobs, and consume media. It is essential that these systems be built from the ground up with freedom of speech and expression in mind, and that U.S. government policy does not interfere with that objective. … The distribution and diffusion of American technology will stop our strategic rivals from making our allies dependent on foreign adversary technology.”

For Chinese audiences, this is evidence that the democratic AI governance narrative under Trump is more about competitive advantage than principle.

Distillation Accusations

The distillation accusations landed in China as a bad-faith political attack dressed up as a security concern. A framing that came up repeatedly was ‘the thief crying thief’ (贼喊捉贼). Many outlets, like Guancha’s 关心 Guanxin column, say Anthropic trained its models on internet data scraped without authorization, then accused Chinese companies of “distillation” and framed it as a foreign attack requiring government intervention. 36Kr made the further point this was a lobbying document timed to coincide with the Pentagon negotiations, an attempt to invoke the China threat to win a contract dispute.

Guanxin made a related point that’s been gaining traction across Chinese tech commentary, which is that Anthropic inadvertently made the strongest possible case for open-source AI. Anthropic claims it could identify individual researchers at Chinese labs from API metadata, by tracking query patterns down to specific employers.

“Anthropic, intending to attack its competitors, inadvertently became the most powerful advertisement for open-source AI. Their actions demonstrated to everyone that under the architecture of closed-source AI services, your privacy, your autonomy, and your right to know are all unprotected. When a company can monitor, judge, and punish you at any time in the name of ‘security,’ so-called ‘trust’ is no longer a virtue, but a risk.”

This argument feels a bit presumptuous, since open-source models have their own API businesses, which offer providers comparable visibility into customer workflows. But perhaps the essential claim is that open-source models can be self-hosted, run locally, with no API calls to the original developer at all.

Share

心智观察所 Xinzhi Observatory, another Guancha column, voiced a more nuanced opinion. It argues that Anthropic’s attitudes towards both China and the Pentagon are consistent with the company’s longstanding worldview.

“[Amodei’s] core argument is not ‘a particular country is dangerous’ but ‘highly capable AI is inherently dangerous.’ In his view, regardless of whose hands a model falls into, the absence of constraints is sufficient for it to be weaponized for mass surveillance or autonomous weapons systems. The intellectual roots of this position can be traced to the influence of effective altruism and long-termism. The logic runs: once AI capabilities cross a certain threshold, they may produce structural risks — and constraints must therefore be built in before deployment. [...] In invoking national security language in its accusations against Chinese companies, Anthropic has, objectively speaking, participated in America’s tech-competition narrative toward China. But its fundamental starting point is concern about ‘capability proliferation,’ not hostility toward any particular nation. It can criticize Chinese companies for distillation, and it can also refuse to grant the U.S. military ‘blanket authorization’ for military use cases. It draws red lines in both directions.”

The Dissolution of US AI Governance

Putting Anthropic aside for the moment, Chinese commentary is drawing some broader structural conclusions about what this episode reveals about the US’s approach to AI governance.

The most common read, unsurprisingly, is that the Washington-Silicon Valley rift exposes a fundamental instability in the American AI ecosystem. State-affiliated general news outlet 澎湃 framed this primarily as a Silicon Valley vs. Washington D.C. story, noting that 550+ Google and OpenAI employees signed an open letter supporting Anthropic. TMTPost 钛媒体, a leading business and tech news, goes a step further in predicting the end of the Washington-Silicon Valley alliance altogether:

“This marks the moment when the covert power struggle between Washington and Silicon Valley — over AI control, the limits of military applications, and tech ethics — finally dropped all pretense and broke into open, no-holds-barred confrontation.”

China, by contrast, has already resolved this question — at least according to many Chinese observers. There was never a pretense that commercial AI companies could set their own limits on military use. The US is discovering messily and publicly what China settled structurally years ago, which is that frontier AI is a powerful technology with deeply dual-use implications, not solely a commercial product with obvious ethical opt-outs. As the aforementioned TMTPost piece puts it:

“[…] idealists like Anthropic who try to walk a tightrope between commerce and ethics are destined to be under the wheels of power […] In the track of artificial general intelligence (AGI), there has never been a so-called ‘neutral zone.’ In the coming months, the battle between Washington and Silicon Valley over model control, underlying values and business interests will surely usher in more intense pains. The final outcome of this game may have a more profound impact on the future of humanity and AI than any iteration of technical parameters.”

Domestic surveillance in China is a de facto assumption, with all companies required to surrender user data to the government if requested. That being said, Chinese analysts have not reached a consensus on Anthropic’s other red line: autonomous drone strikes. Back in February 2025, Peking University Professor Zhu Qichao 朱启超 contributed an op-ed about AI and the ethics of autonomous weapons to the People’s Daily, the official newspaper of the Chinese Communist Party’s Central Committee. The publication of analytical writings like this in top state media outlets is a good indication that for decision-makers in Beijing, this is a topic worthy of further study and debate rather than a settled matter. Zhu wrote:

“When an AI system malfunctions or makes a flawed decision, should it be treated as an independent entity bearing responsibility? Or should it be treated as a tool, with human operators bearing all or part of the liability? The complexity of this accountability question lies not only at the technical level but also at the ethical and legal levels. On one hand, although AI systems are capable of autonomous decision-making, their decisions remain constrained by human-designed programs and algorithms — meaning their liability cannot be entirely separated from human responsibility. On the other hand, AI systems may in some circumstances exceed the parameters humans have set and act on independent judgments; how to define accountability in those cases has become a persistent challenge in arms control. […]

As AI is applied ever more deeply to military contexts, the human role within combat systems is shifting — from the traditional ‘human-in-the-loop’ model toward ‘human-on-the-loop,’ with humans evolving from direct operators inside the system to external supervisors monitoring it from without. This transition, however, raises new questions of its own. Ensuring that AI weapons systems continue to adhere to human ethics and values when operating independently represents one of the most significant challenges currently confronting the field of AI weapons development.”

Subscribe now

For many in China who look to the US as a place where a safety-focused company could resist state capture, where Anthropic’s model of principled refusal was even theoretically possible, that idea has now taken a big hit. Weijin Research 未尽研究, an independent analysis firm, argued in a piece that came out before this dispute, “Anthropic’s safe-first principle functioned not only as a moral standard-bearer but also as a powerful commercial moat — one that proved especially effective in enterprise and government markets.” Quoting ’s commentary on the Pentagon’s decisions, Weijin Research asserted that the situation is a “warning for the entrepreneurship ecosystem and talent flows […] under this political environment, is any tech company truly safe?”

Taiwanese Perspectives

Does the threat of falling behind China justify tabling ethical questions about military AI? Some Taiwanese defense analysts think the world would be better off if Anthropic chose to work within the system.

Pei-Shiue Hsieh 謝沛學 at Taiwan’s Institute for National Defense and Security Research (INDSR) writes:

“Non-democratic regimes possess an ‘asymmetric advantage’ in the military application of AI. The standoff between Anthropic and the U.S. Department of War over ‘lethal autonomous weapons’ reflects an uncomfortable truth: setting aside technological and economic capabilities, democracies have inherent disadvantages and limitations in the military application of artificial intelligence — particularly in the development of ‘lethal autonomous weapons.’ The ‘don’t be evil’ principle may occupy the moral high ground, but it only has influence over policymakers and corporations in democratic countries; politicians in non-democratic states are entirely unconstrained by it.

This is analogous to how the restrictions the Intermediate-Range Nuclear Forces Treaty (INF Treaty) imposed on the United States allowed China — which refused to join the treaty — to build up an advantage in intermediate-range ballistic missiles and area-denial capabilities in the Indo-Pacific region.

Let us posit a scenario here: Anthropic’s resistance succeeds and triggers a chain reaction, Silicon Valley’s tech mainstream reverts to its stance of withdrawing from defense contracts, and the U.S. military’s military AI development is severely impeded as a result. Meanwhile, China is able to integrate AI into all manner of military R&D without restraint, ultimately achieving an overwhelming advantage in military AI — particularly in ‘lethal autonomous weapons.’ Would such a world be safer?”

Meanwhile, other analysts lamented the emerging race-to-the-bottom dynamic. One writeup called the dispute “the AI industry’s first coming-of-age ceremony” (成年禮). Another author, “Future Lin,” wrote on Substack:

This is a decisive moment for AI governance, not a neutral policy debate. The core issue is not whether Anthropic should concede, but rather: “When governments have the weapon to designate tech companies as national security threats, who dares to say no to the military?” Taiwan’s AI industry is not a bystander, because once this logic becomes entrenched, the ethical accountability mechanism of the global AI supply chain will be fundamentally shaken.

…

For decades, the U.S. tech industry’s advantage has partly stemmed from its relatively independent operating logic — the government can procure, but it cannot make unlimited demands. This boundary is one of America’s invisible assets for attracting top global AI talent: you can start an AI safety company here without worrying that the government will forcibly repurpose your technology for something you consider harmful.

That premise has now been compromised.

The implications for Taiwan are more direct: many Taiwanese AI startups have business plans that include the U.S. market and U.S. government contracts. In this new environment, the assessment of “whether you can land a U.S. government contract” must incorporate a new dimension — if you have ethical boundaries, and those lines conflict with what the government demands, what consequences are you prepared to bear? The more fundamental question is this: if the ethical standards of the global AI supply chain are dictated by the government agencies with the greatest purchasing power, where is the market space for the very concept of “AI safety”?

Share

Social Media

Living next to an authoritarian superpower that faces no such internal friction, some Taiwanese commentators see Anthropic’s ethical stand as a luxury democracies can’t afford.

On Threads:

This whole thing is obviously just Anthropic being idiots (當小87). The company isn’t like Google, with data centers and energy infrastructure spread all over the world, and it doesn’t have the ability to develop its own hardware either. Under those conditions, its bargaining position was already weak to begin with. Because its supply chain has to follow the U.S. government and the military anyway, it basically has zero leverage to pursue some “tech-lefty” agenda. Google employees can afford to play some progressive political games because the company’s fundamentals are strong enough to support that. Anthropic doesn’t have that luxury at all — it’s basically just making more investors want to pull their money out.

From the PTT Stocks board:

Is it possible for your enemy, China, to do such a thing?

If the PLA were to use Deepseek, would Liang Wenfeng dare to tell them, “You can only use Deepseek for XXX, not OOO.”

Therefore, the same pressure forcing Anthropic’s hand reflects genuine urgency inside the Pentagon about closing the gap with China — particularly in drone warfare, where the quality of Chinese drones has spooked Washington. If a Taiwan contingency ever materializes, the DoW at least seems serious about not showing up to that fight with inferior AI.

It’s been a big week for Chinese AI video.

ByteDance’s Seedance 2.0 and Kuaishou’s Kling 3.0 are going viral. Seedance probably takes the cake, being promoted for smoother motion, stronger scene coherence, and watermark-free downloads.

My favorite trend so far: people making videos of themselves cooking Lebron James.

On the other hand, the Cyberspace Administration (CAC) just announced its pre-Lunar New Year crackdown by removing more than 540k pieces of illicit AI content on Douyin, Wechat, and Weibo and taken action against over 13k accounts, making a statement that it’s stepping up enforcement of unlabeled or “garbage” AI-generated content.

China’s AI video ecosystem is moving at hyper-speed, while the CCP, typically the world’s most adept and overbearing content regulator, is scrambling to tighten control in real time. This is a bit puzzling, since China already has one of the most ambitious frameworks in the world for labeling and managing AI-generated content. If the rules were put in place for a moment like this, why does it feel like they’re not working?

Today:

• China’s new rules for labeling AI-generated content

• How those rules are being poorly enforced

• The domestic incentives behind this poor enforcement, including economic growth priorities

• The geopolitical incentives, including the use of AI to spread pro-China and anti-US propaganda

Part two will give a broader industry overview of how Chinese models compare to Western ones, which companies are leading and targeting the Global South, and why video generation poses different training, open-source, and copyright implications for the Chinese ecosystem.

The Rules

China was the first country to introduce rules specifically for synthetic media—the Deep Synthesis Provisions (互联网信息服务深度合成管理规定), which took effect in the dark ages of 2022.

Fast forward to March 2025: the Cyberspace Administration of China released the Measures for Labeling of AI-Generated Synthetic Content (关于印发《人工智能生成合成内容标识办法》的通知). These rules require platforms to detect when content is confirmed, likely, or even just suspected to be AI-generated, and to attach both visible labels and embedded metadata by September 1st, 2025.

The accompanying national standard GB 45438-2025 spells out what those labels have to look and sound like:

• Images: visible text label covering at least 5% of the shortest side length

• Audio: spoken label at 120–160 wpm, or a Morse-code-style tone representing “AI”

• Video: clear labels shown at both the beginning and end of the clip

• Metadata: persistent, machine-readable tagging embedded in the file

China is essentially the only major country to enact a preemptive, upstream requirement that all AI-generated video be labeled, watermarked, and traceable across the stack.

Most other countries are still operating on a reactive, harm-based model. South Korea and the UK criminalize certain uses of deepfakes if they cause harm (mainly sexual or reputational). Singapore’s Elections (Integrity of Online Advertising) Amendment Act bans deepfakes in election advertising, with penalties up to five years in prison. Russia’s Federal Law No. 32-FZ tightens criminal-code penalties for the dissemination of “false information” about the armed forces, including deepfakes or synthetic media used to “misrepresent” military operations in Ukraine. The US has little beyond a patchwork of state rules for political ads and intimate-image abuse (e.g., the TAKE IT DOWN Act), though Google and Sora voluntarily watermark their content.

The closest comparison to China is the EU’s AI Act, which mandates AI content labeling as of August 2025. However, decentralized enforcement across member states with varying regulatory frameworks will lead to fragmented enforcement across the EU, likely limiting consistent implementation.

Subscribe now

Bending the Rules

China’s rules were supposed to take effect on September 1st, 2025. Yet if you go on Chinese social media right now, you’ll find an abundance of AI-generated videos with no labels at all.

You’re supposed to see something like this label in the corner of every piece of AI-generated content:

You might also see something like this at the bottom, which reads, “This content could be AI-generated”:

Content creators can also choose to self-report when they’ve used AI:

The government’s hope is that platforms will also detect this automatically through the video’s metadata. But in reality, only Chinese AI companies are beholden to including compatible identification information in their metadata, and Sora- and Veo-generated clips are quite popular on Chinese social media, usually producing the funniest content, since many Chinese models tend to avoid being subversive. Therefore, once exported and uploaded, platforms often cannot reliably determine whether a clip is synthetic. And with Chinese social media platforms locked in fierce competition, both with each other and the Western market, none wants to be the strictest enforcer while others let content flow freely.

As mentioned in the intro, the CAC says it removed more than 540k pieces of illegal AI content and acted against 13k accounts. That sounds big until you compare it to the roughly 34 million videos posted on TikTok or the 100 million throughout Meta just in one day! At that volume, 540,000 clips amount to well under 1% of a single day’s output — mathematically, a drop in the bucket.

Perhaps I’m being harsh. This announcement is meant more as a signal that the CAC is taking regulating AI content seriously rather than trying to single-handedly de-slopify the platforms. But at this scale, it risks sending the opposite message, making the CAC look less in control, not more, especially when companies like ByteDance are doing things like explicitly promoting watermark-free videos.

There is also a plethora of content online without any identification whatsoever:

You’ll commonly see AI-generated ancient towns, places in China that are subtly distorted, and landscapes that look believable enough to fool an untrained eye (all unlabeled). My mom sometimes sends me clips and asks me, “Have you been here?”

I have not.

There’s also a trend I’ve noticed of half-real, half-AI-generated content, which further blurs the precedent of whether it should be labeled. (This aligns with Part 2’s discussion of how Chinese companies are succeeding through AI-enhanced editing tools like CapCut rather than purely generative text-to-video models.) Scroll through RedNote and you’ll instantly find clips where the hill or building in the shot is clearly real footage, but the clouds shift and morph in a way that’s unmistakably artificial:

Source.

Domestic Incentives

What gives?

As lamented in our Chinese AI in 2025, Wrapped:

“AI-assisted and -generated content is now so much more pervasive online than nine months ago, whether on global platforms or on the Chinese internet. It’s time to ask: what was the point of labelling as policy? Is it to actually protect users from misinformation and engender trust, or is it just a stopgap measure that lets platforms evade responsibility? What kinds of AI usage merit which kinds of mandated disclosures?”

One rationale is that China is tightening controls on misinformation while simultaneously pushing generative video as a pillar of its AI+ strategy for digital commerce and economic revitalization — a fine line to walk.

Subscribe now

State media constantly praises AI-generated content’s potential as a new economic stimulant. The CCP is partially betting on AI video as a core part of e-commerce, short-form entertainment, and platform growth. Micro-dramas, digital-human product demos, and auto-generated marketing clips are widespread.

The government itself is generating slop. Chinese broadcasters have rolled out AI avatar news anchors, and the CCP has readily broadcast AI content through official CCTV programming; for instance, Poems of Timeless Acclaim 千秋诗颂, which uses the “CCTV Listening Media model” to transform classical poems into ink-wash animations.

Geopolitical Incentives

Those same tactics also underpin China’s AI push in the Global South, which accounts for a large share of its AI video user base. If its video-generation ecosystem restricts content too heavily, it’ll struggle to attract new users outside of China who expect the creative freedom offered by international competitors.

There have been reported cases of Chinese state actors experimenting with deepfake presenters to push pro-China/anti-US narratives in local languages, while others have promoted multilingual “AI anchors” (数字人主播) that can run 24/7 and auto-generate product videos or livestream scripts for cross-border commerce. (I was, however, only able to find a small number of documented cases of such practices, with no sign of widespread adoption.)

Beijing also doesn’t mind when the content helps give them a boost.

AI-generated clips that mock foreign rivals tend to circulate without much issue. This year, there’s been a wave of videos about US H-1B visa frustrations, military action in Venezuela, or overweight American factory workers struggling to revive manufacturing in the wake of Trump’s tariffs; content that flatters China’s self-image and reinforces familiar state narratives.

^A college student finally saves up enough money for her H1B visa. Source.

^‘Live-cam footage’ of US soldiers arresting Maduro. Source.

I’m not doubtful the CCP’s system is effective at blocking content it views as potentially destabilizing. You won’t find Mao Zedong and Chiang Kai-shek sharing a beer or setting the Olympic pole vaulting world record (as I routinely do on my Instagram feed). And you’ll rarely see political dissent of any kind. But for the remaining slop, Beijing seems to tolerate a certain amount of slippage, especially if it can help grow the ecosystem and serve its other interests.

Vibe

An aesthetic observation to close.

After spending too much time on Chinese social media “researching” this article, I’ve come away thinking that AI video is seen in a more optimistic light overall than the doomerish perspective you’ll find pervading many of the comments on Western platforms.

For instance, it’s quite common to find depictions of the future, often cyberpunk-esque, that fuse ancient Chinese civilization with radical future cityscapes. I’m sure this exists somewhere on Western social media beyond my filter bubbles, but it doesn’t seem as prominent.

This sinofuturistic temperament — the willingness to embrace rapid technological progress in a utopian or at least neutral register rather than an inherently dystopian one — imbues the trajectory of this technology with a greater sense of wonder. It foments an eagerness to see what worlds might be built, rather than fixating exclusively on the reality it might subsume.

Don’t worry, there’s also plenty of brainrot slop.

*Bonus clip: The ChinaTalk team tries out Kuaishou Kling’s image-to-video feature at NeurIPS 2025:

In terms of international expansion, Chinese firms are way ahead of the American competition. Chinese companies have worked out Autonomous Vehicle (AV) deployment deals with more than thirteen countries. The US: two. Chinese companies are also exporting something closer to a full autonomy stack — vehicles bundled with cloud services, AI traffic management systems, and road sensors.

There’s also the supply chain. Unlike frontier AI models, where US export controls on Hopper and Blackwell GPUs have genuinely constrained China’s progress, AVs operate in a different hardware regime. Here, the leverage between the US and China is more evenly matched, and in some cases, inverted.

Today’s Content:

1. AVs in the US and China

2. The International AV market

3. Who has leverage in the AV supply chain

Subscribe now

At times, this piece reads like a typical “US vs China” article, but in fact we’re seeing more of a “co-opetition” dynamic highlighted in the AI industry. In fact, the perhaps more interesting aspect is how the line between “Chinese AV company” and “US AV company” blurs in practice. Chinese AVs use NVIDIA chips, Waymo uses Chinese-made Zeekr vehicles, and Uber and Lyft partner with Chinese AV firms internationally, not to mention critical minerals. The industries are too tangled for neat distinctions… but let’s try to untangle them anyway.

1. AVs in the US and China

The US has Waymo. China has its “big three”: Baidu’s Apollo Go, WeRide, and Pony.ai (whose founders ChinaTalk interviewed last year).

There are other potential players in the US, like Amazon’s Zoox and Tesla (RIP GM’s Cruise). But right now, Waymo is the only American company operating a scaled, paid Level-4 robotaxi service, which enables vehicles to handle all driving tasks within specific operating zones. China also has BYD and Xiaomi with L2 driving features (who could transition to L4 soon), and many more robovan, robus, robodelivery, and robotruck companies on track for L4 deployment.

In aggregate terms, China appears to have the edge in overall deployment. An analysis by SCSP suggests Chinese autonomous-vehicle operators have collectively logged roughly 149 million autonomous miles, compared to around 106 million miles for US firms — a roughly 1.4 to 1 advantage.1

But mileage comparisons are limited. Companies report different levels of autonomy, mix supervised and driverless miles, and disclose data unevenly across jurisdictions. Ridership is a different way to look at it, where China has completed ~30 million rides, versus ~20 million for the US (Breakdown in Appendix 1).

Different Services

Ridership itself misses a big part of the story, because China’s AV industry extends beyond passenger ride-hailing. By the end of 2024, more than 6,000 driverless delivery vehicles were reportedly operating across 100+ city zones. Companies such as Neolix, Zelos, Meituan, JD Logistics, and Alibaba’s Cainiao are actively piloting or scaling operations for shipping, food delivery, and street-cleaning vehicles.

The US, by contrast, doesn’t yet have road-going (as opposed to sidewalk-going) driverless AVs deployed, with companies like Nuro still limited to pilots and R&D fleets. They do have an estimated 3,000 to 5,000 sidewalk AVs, but China has many of these also, and they are a much simpler technology.2

China also seems to be leading in autonomous trucking.

Different Environments

A common critique is that China’s pilot zones create an artificial environment, so bragging about its safety record or miles driven is like bragging about avoiding accidents while driving a golf cart around a country club. I think this is wrong.

For context: China has pioneered an ambitious pilot zone strategy, reshaping large portions of cities to accommodate AVs through vehicle-road-cloud integration (车路云一体化). Intersections broadcast signal timing, cameras extend line-of-sight, and cloud systems coordinate traffic flows. These zones feature highly visible stop signs, walkways, and pedestrians, often exclude unpredictable bike lanes and scooters, and were initially launched in lower-density areas.

But Waymo can’t drive everywhere in its pilot cities either. Wuhan’s pilot zones started small but now blanket the entire city, with AVs driving alongside cyclists, scooterers, and jaywalkers.

^Guy rides an Apollo Go vehicle in Wuhan, one of the most ambitious AV cities in China. Also, note that Chinese AVs seem to have much more expansive user experience features, like an AI assistant that can roll down the window or give you a back massage.

There’s also a more forgiving regulatory structure. China’s centralized system somewhat mitigates complications like interstate travel. The US counter is that bullish states can move ahead without federal approval, but the downside is that Waymo burns significant resources negotiating bespoke city-by-city agreements domestically, diverting time from international expansion and raising legal barriers for new companies entering the US market. Chinese AV companies also have to work out separate agreements city-to-city, but with a federal mandate and more lenient legal code, this is still less of a headache than in the US.

Differences In Public Opinion

The survey evidence we have (imperfect and limited as it is) points in a fairly consistent direction: Chinese consumers appear meaningfully more comfortable with autonomous vehicles than their American counterparts. Across multiple studies, acceptance levels in China typically fall somewhere in the 50-80 percent range, while US figures tend to cluster closer to 25-40 percent.

High-profile incidents have, however, exposed real limits to public tolerance. A widely reported crash involving a Xiaomi SU7 that killed three students triggered a wave of public backlash and appears to have made regulators more cautious about how quickly AVs are rolled out.

Share

There’s also a growing undercurrent of economic resistance. In Wuhan, often described as China’s most AV-forward city, ride-hailing drivers have repeatedly complained (you might even say protested) that autonomous vehicles threaten to displace what is, for many, their primary source of income. Those concerns carry extra weight in a weak labor market, where unemployment remains high and alternative opportunities are scarce.

China is reported to have around 38 million truck drivers. In the US, it’s roughly 2.2 million, and the industry faces labor shortages. China also has a larger ride-hailing and food-delivery ecosystem, meaning that when you add it all up, China has nearly twice the per capita rate of workers in occupations at risk of displacement (36.8 vs 20.0 per 1,000 people) [Breakdown in Appendix 2].

2. The International AV Market

The divergence between US and Chinese approaches to autonomous vehicles becomes even more pronounced in their international expansion strategies.

US Companies (i.e., just Waymo currently) have deals with:

1. USA

2. UK

3. Japan

4. Possibly Australia

Chinese companies have deals with:

1. China

2. UK

3. UAE

4. Saudi Arabia

5. Singapore

6. Hong Kong

7. South Korea

8. Germany

9. France

10. Spain

11. Switzerland

12. Belgium

13. Luxembourg

The international lead for Chinese companies is larger than it appears, because countries already integrated into China’s Digital Silk Road infrastructure are natural targets for the next wave of expansion. Egypt, for example, is attempting to ditch Cairo by building the New Capitol, with extensive Chinese infrastructure support for “smart city features,” the kind of project that sounds ripe for AV deployment. Similar rumblings have been true for Oman. WeRide also struck a partnership with Grab, the biggest ride-hailing app in Southeast Asia, which positions it well for expansion into the entire subcontinent.

*Note: What counts as an ‘agreement’ is broad. France’s agreement with WeRide for robobus deployments at Roland-Garros isn’t blanketing Paris with robotaxis, but it still shows the buds of cooperation, such as the subsequent deal between WeRide and Renault to deploy Robobuses in France’s Drôme region. Others, notably in the Middle East, are signalling the launch of full-scale commercial operations with hundreds of vehicles and city-wide infrastructure.

Subscribe now

Middle East

China has the clearest advantage in the Middle East.

What makes the Middle East such a natural fit is that many governments there can reconfigure the built environment and the regulatory environment in parallel to accommodate AVs. The US strategy, thus far, seems to be to drop AVs into existing roads and regulatory systems and hope for the best.

For example, Saudi Arabia signed a deal with WeRide in 2024 to deploy robotaxis across the Neom megaproject. With help from the Chinese, the planned city is being built with dedicated infrastructure for autonomous vehicles. The UAE is doing similar things. WeRide has reportedly accumulated 1 million km of operational mileage in Abu Dhabi already.

Some of these supply-side projects sound overly ambitious (not off-brand for Chinese infrastructure projects), but the crucial difference from Chinese domestic projects is that Chinese companies won’t primarily bear the financial burden if these Middle Eastern ventures turn out to be total disasters. The host countries are paying Chinese companies to provide infrastructure, meaning China profits from its construction expertise regardless of whether demand materializes. If Neom ends up as a half-built, empty city in the middle of the desert with hundreds of AVs and no one to use them, that’s a sunk cost primarily for the UAE and Saudi Arabia.

China would still incur losses from producing the vehicles themselves, but if they only provide the software and key components, like LiDAR, while allowing local companies to supply the base vehicles, losses would likely be limited to $10-20k per vehicle. These vehicles could also be reclaimed and redeployed elsewhere.

Europe

Europe is a more surprising case than the Middle East. Despite the EU’s openly wary stance toward Chinese vehicles — as evidenced by their anti-subsidy probes and tariffs on Chinese EVs — Chinese AV companies are making significant inroads ahead of American competitors.

Is Europe following the same trajectory with respect to Chinese EVs, where regulators stepped in only after there was significant market penetration?

From all my interviews on this topic, the biggest open question people had is what will Europe do next? GDPR already imposes strict requirements on data collection and processing, while the less-discussed Cyber Resilience Act mandates cybersecurity standards for connected products. Both could be used to block Chinese AVs.

Share

The precedent is already set. The US has effectively banned Chinese EVs through the Connected Vehicle Rule, which prohibits vehicles with Chinese or Russian software due to national security concerns about data collection and potential infrastructure mapping. However, China has recently responded with new guidelines exempting foreign-collected automotive data from security assessments, primarily an attempt to assuage European data privacy concerns about Chinese vehicles.

Chinese companies are partnering with local European automakers (like Stellantis) for the base vehicle while supplying the AV-specific infrastructure and software themselves. This approach could help circumvent EU tariffs on Chinese EVs — which would otherwise apply to Chinese robotaxis if they use Chinese EVs as their base vehicle — and positions the arrangement as a joint venture rather than simply dumping Chinese vehicles onto European streets.

3. Who has leverage in the AV supply chain?

China’s Leverage

LiDAR uses laser pulses to create detailed 3D maps of a vehicle’s surroundings, measuring distances to objects. It’s the spinning sensor thingy typically mounted on top of autonomous vehicles. China controls ~90% of the global LiDAR market. Firms like Hesai, RoboSense, Huawei, and Seyond account for the bulk of automotive-grade LiDAR shipments.

Waymo, importantly, produces its own LiDAR in-house, so China can’t cut it off. But most of the AV industry lacks that level of vertical integration, and because Waymo’s LiDAR is proprietary, new companies entering the market will likely be beholden to Chinese suppliers. (An AI analogy is Google’s TPUs for AI models; they can produce these for themselves, but a new entrant into the AI game would likely be reliant on NVIDIA’s GPUs.)

Batteries are another potential choke point. Most robotaxis are electric, which pulls AVs directly into the EV battery supply chain that China dominates end-to-end. CATL and BYD together account for over half of global EV battery installations, with companies like Tesla, BMW, Ford, Volkswagen, and Toyota all using them. Even Waymo’s next-generation robotaxis use Chinese-made Zeekr vehicles powered by batteries from CATL.

Interestingly, CATL was designated as a “Chinese military company” by the Biden-era DoD and added to the blacklist for government or military usage. This blacklisting doesn’t apply to commercial vehicles… yet.

These advantages translate directly into cost. Public estimates routinely put Waymo’s current robotaxis at roughly $130k-$150k per vehicle, once sensors, compute, and the base car are included. It is purported that they spend $40-50k just on sensors (like LiDAR), since they are paying a premium to produce it themselves. Chinese robotaxi platforms, by contrast, are cited at $30k-$50k all-in, due to state subsidies and the preexisting car manufacturing prowess.

US Leverage

If China were to withhold access to LiDAR or batteries, how could the US respond?

Most directly through NVIDIA chips, but not the standard Hopper and Blackwell series GPUs used in AI data centers. Unlike data center GPUs designed for training, AVs use automotive-grade systems-on-chip (SoCs) like NVIDIA DRIVE Orin and Thor — integrated platforms that combine CPU, GPU, and dedicated neural network accelerators optimized for real-time inference, safety certification, and lower power consumption. Baidu uses dual Nvidia Orin X chips, Pony.ai uses four Orin chips, and WeRide recently deployed Nvidia’s Thor platform.

Subscribe now

However, the technical barriers differ significantly between AI and AV chips. For AI GPUs, cutting-edge nodes (2-5nm with EUV lithography) are essential to achieve the compute density required for training workloads, but autonomous driving chips face lower node requirements. Current AV chips like Nvidia’s Orin operate on relatively mature ~8nm processes, since AV workloads prioritize deterministic latency, power efficiency, safety certification, and software integration over just raw compute density. This means Chinese domestic foundries like SMIC could theoretically produce decent AV chips on 7-12nm nodes without accessing advanced lithography equipment, whereas comparable AI training chips would require the cutting-edge processes that remain out of reach.

I don’t want to underplay it. The upcoming NVIDIA Thor chips use 4-5nm nodes that Chinese companies cannot fabricate through TSMC. And it would be a real headache for Baidu, WeRide, and Pony.ai to switch to domestic alternatives, as this also means switching much of the downstream software ecosystem. But unlike AI training, I also don’t want to underplay that chips aren’t everything for AV infrastructure.

US export restrictions currently limit TSMC to fabricating AI-related chips at 7nm or larger nodes for Chinese customers. AV chips ostensibly fall under this distinction. Several leading Chinese AV chips therefore operate at or above this threshold: Horizon Robotics’ Journey 6P chip, Huawei’s HiSilicon Ascend series (used in its MDC platform), and Black Sesame’s A1000 chip. These chips are proving, at minimum, viable: Black Sesame’s A1000 powers mass-production vehicles from Geely’s Lynk & Co, Hycan, and other major Chinese automakers, while Horizon controls 49% of China’s self-driving chip market and claims more vehicles with Navigate on Autopilot-style features use its chips than NVIDIA’s.

So, NVIDIA has leverage in AV chips, but lacks the stranglehold it enjoys over AI GPUs. I do still believe their chip advantage outweighs the leverage Chinese companies hold with LiDAR and batteries, since China remains fundamentally bottlenecked without access to EUV lithography and other crucial semiconductor manufacturing equipment, whereas the US has demonstrated it can produce LiDAR and batteries when needed, albeit less efficiently and at higher cost than China. However, rare earths and critical minerals add another layer of interdependence, since elements like neodymium and gallium are essential for LiDAR systems and electric motors. Everything is quite intertwined at both the front end and back end of the stack, meaning, for now, both American and Chinese vehicles are reliant on each other.

Conclusion

China’s AV sector is performing strongly — controlling key parts of the supply chain, dominating international expansion, and scaling up deployment domestically. But here’s a puzzle: why did both Pony.ai and WeRide experience brutal post-IPO crashes? Pony.ai fell from its debut price of $15 to $4.18 before recovering to around $16-17, while WeRide plummeted from $15.50 to around $9-10. Their valuations of approximately $7 billion and $3 billion pale in comparison to Waymo’s estimated $110+ billion valuation.

I suspect these conditions can coexist. China’s unique political economy has proven it can (1) dominate a global sector without (2) guaranteeing it is financially lucrative. A similar dynamic has played out with EVs. China produces 70% of global EV output, yet most Chinese EV makers operate at a loss. State-backed capacity buildout created severe overcapacity and price wars, with automaker margins falling from 5.0% in 2023 to 4.4% in 2024 despite surging volumes.

The brutally competitive domestic market may explain why these AV companies are racing overseas. If you strike a deal with the UAE and are the only AV company able to operate there, there is some room to breathe away from the vicious competition within China.

Want to chat about global AV competition? Reach out to me at nick@chinatalk.media

Share

Appendix 1: Estimate of Total Rides

• Waymo (US)

  • Widely cited reporting places Waymo at ~20 million cumulative paid robotaxi rides as of early 2026 (Financial Times)

• Baidu Apollo Go (China)

  • Public disclosures and secondary reporting put Apollo Go at ~17 million cumulative rides by late 2025 (Financial Times, Baidu Press Release)

• WeRide (China)

• 1000 robotaxies (WeRide website)

• Conservative implication:

  • 10 trips/vehicle/day → 1,000 × 10 × 365 ≈ 3.7M rides/year

  • 25 trips/vehicle/day → 1,000 × 25 × 365 ≈ 9 M rides/year

• Pony.ai (China)

  • Chinese-language disclosures report ~900-1,000 robotaxis with ~15-23 orders per vehicle per day

  • Implied annual volume:

    • 15 trips/day → ~5M rides/year

    • 23 trips/day → ~8M rides/year

• Aggregate implication (best guess)

  • Apollo Go (cumulative): ~17M

  • WeRide (implied cumulative): ~3.7-9M

  • Pony.ai (implied cumulative): ~5-8M

  • China total: ~25.7-34M cumulative robotaxi rides

  • U.S. total: ~20M (just Waymo)

There are other small contributors on both sides, but I think these effectively cancel each other out, since none exceed 1 million rides.

Appendix 2: Estimate of Non-Passenger Vehicles

China had 7.48 million certified ride-hailing drivers in 2024, and food-delivery platforms alone support millions more workers. Meituan alone reported 3.36 million average monthly active delivery riders in 2024, many of them relying on these jobs as their primary source of income rather than a side hustle, whereas that number is around 37% of US gig workers. I estimate the Chinese total is something like 13-14 million for food plus ride-hailing services (plus the 38 million truck drivers).

The US has approximately 3-6 million gig workers in these sectors: ~2-2.2 million ridehailing drivers (Uber had 8.8 million drivers globally in Q2 2025 with approximately 1-1.5 million in the US based on historical data) and ~2.5-3 million food delivery workers. DoorDash had 8 million dashers in 2024, but I’ll also discount this because many of them are part-time workers. 72% of DoorDash drivers work 4 or fewer hours per week.

In total, I estimate:

China: (13.5 + 38 million) ÷ 1,400 million (population) = 36.8 per 1,000 people

United States: (4.5 + 2.2 million) ÷ 335 million (population) = 20.0 per 1,000 people

Happy New Year! ChinaTalk is kicking off 2026 with an audience survey. The link is here. Please fill it out — your feedback is important to us! ~Lily 🌸

We just dropped a Second Breakfast Venezuela emergency podcast for a breakdown of the tactical, strategic, and legal implications of abducting Maduro (listen here). ChinaTalk’s below explores the China angles to this story.

Statements from the Chinese government on the US actions in Venezuela were predictably critical. China’s Foreign Ministry condemned the operation as a violation of international law and the UN Charter, called for the safety and immediate release of Maduro and his wife, and accused Washington of acting like a “world judge” and a “unilateral bully.” Beijing also backed an emergency meeting of the United Nations Security Council, during which they reprimanded the US on their standard grounds of sovereignty, non-interference, and opposition to hegemony.

But beyond official statements, scholars, policy analysts, and online netizens have offered a wide range of interpretations that shed light on how Chinese audiences view US power, international law, and the implications for Taiwan.

Spheres of Influence

Chinese commentators have quickly embraced the neologism “唐罗主义” tángluó zhǔyì, a wordplay riff on Trump’s “Donroe Doctrine,” to frame the US move.

Niu Haibin (牛海彬), director of the Latin America research center at the Shanghai Institutes for International Studies, argued that while oil and sanctions matter, they are secondary: “The main objective is reflected in its new National Security Strategy, which is to rebuild US hegemony in the Western Hemisphere.”

Wang Yiwei (王义桅), the director of the International Affairs Institute at Renmin University, described the operation as evidence that the US is willing to overthrow governments it deems unfriendly to intimidate the region and reassert imperial control.

A key distinction scholars point out is that these actions violate the UN Charter and international law, which matters for Taiwan.

Beijing’s position rests in part on the ambiguity of UN General Assembly Resolution 2758 (1971), which recognized the People’s Republic of China as the sole legitimate representative of “China” at the United Nations and excluded the Republic of China (Taiwan) but did not explicitly resolve Taiwan’s legal status or state that Taiwan is part of the PRC. Chinese analysts invoke this ambiguity to argue that cross-strait issues fall outside the scope of international intervention, whereas US actions in Venezuela fall within that scope. Chinese scholars are therefore not arguing that the US’s actions justify carving out their own Monroe-like sphere of influence in the South China Sea or East Asia, but rather that the situations are disanalogous.

On Weibo (Chinese Twitter), some Chinese netizens have openly described the episode as a Taiwan template, arguing that it shows how quickly a great power can act, impose a fait accompli, and only afterward fight over legitimacy:

“The situation in Venezuela gives us an idea for unifying Taiwan: We could launch a special forces operation to capture Lai Ching-te, then immediately announce the takeover of Taiwan, change the identity cards the same day, and achieve a quick victory.” Source.

Others, however, pushed back against drawing a direct analogy. Some warned that equating the two cases was strategically reckless, stressing that Beijing claims far stronger legal and historical justification for Taiwan as an internal matter than the US does for intervening in Venezuela:

“Please be reminded that a US military strike against Venezuela would be a serious violation of international law and an act of aggression against a sovereign state. However, any action we take regarding Taiwan is our internal affair, and no other country has the right to interfere. These two situations are not the same, so don’t be misled by certain opinions.” Source.

For many commenters, though, the more salient takeaway was Washington reverting to a colonial or imperial mode of behavior, with netizens invoking histories of Western aggression in Asia and questioning why this intervention is treated differently from Russia’s invasion of Ukraine. In this telling, China’s restraint — whether in Venezuela or Ukraine — is recast as vindication rather than passivity: proof that Beijing, by staying out, avoids exposing the coercive instincts that Western powers reveal when they intervene abroad under ideological pretexts that critics say often mask more material interests, like oil.

On Taiwanese social media, the reaction has been different.

ChinaTalk Taiwan correspondent Lily Ottinger notes that much of the online discussion has focused on military performance. Bloomberg reported — later rewritten with direct quotes by Taiwan’s Central News Agency — that a senior Taiwanese national security official viewed the episode as helpful for deterrence, signaling that President Trump is willing to use force in defense of what he sees as core US interests, and that US forces can overwhelm militaries reliant on Chinese equipment.

From a PTT (Taiwan’s popular Reddit-like forum site) discussion of the comments:

“Chinese radar, Russian missiles, it’s really a joke.”

“It turns out Chinese radar is garbage; air superiority in the Taiwan Strait is basically firmly in Taiwan’s hands.”

“The quality of that Chinese radar makes one wonder if it was bought through Pinduoduo.”

Some Taiwanese legacy media went further. One article by the DPP-leaning Liberty Times (自由時報), titled “The Failure of the China Model,” argued that despite Venezuela being one of China’s closest military partners in South America — operating Chinese-made radar systems, K-8 trainer aircraft, and armored vehicles, and reportedly hosting Chinese military advisers — the operation revealed how little that partnership translated into real defensive capability. China’s “defensive shield” collapsed under pressure, and Beijing’s lack of response reinforced the perception that China is a limited security partner when confronted with US forces.

Subscribe now

Venezuela’s Chinese-supplied radar network failed to detect or deter US aircraft and was quickly neutralized, overwhelmed by superior electronic warfare and precision strikes. But this says only so much about the quality of Chinese weapons. China has not supplied Venezuela with its most advanced systems, and many of the country’s more serious air-defense capabilities — such as surface-to-air missile systems — were sourced from Russia and poorly paintained. Seen this way, the episode could reflect less a failure of Chinese hardware than the limits Beijing has deliberately placed on how far it is willing to militarize partners in the Western Hemisphere.

Venezuela-China Relations

Over the past two decades, Beijing has persuaded a steady stream of Latin American countries — including Costa Rica, Panama, the Dominican Republic, El Salvador, Nicaragua, and Honduras — to switch diplomatic recognition from Taiwan to China, often pairing diplomatic pressure with promises of investment and strategic partnership. Venezuela made that switch much earlier, in 1974. Ties deepened after Hugo Chávez took power in 1998, and when Maduro took power in 2013. Maduro even enrolled his son at Peking University in 2016.

In 2023, China and Venezuela elevated ties to an “all-weather strategic partnership” during a meeting between Maduro and Xi in Beijing, a designation Venezuela shares with only a few other countries like Pakistan and Belarus. But that distinction apparently excluded taking action to defend Venezuela from the US.

The asymmetry in the relationship was visible even at the end. Maduro’s last publicly reported meeting — just hours before his capture — was with a Chinese special envoy sent to reaffirm Beijing’s support. But the meeting was with a relatively low-level Chinese delegation at a moment of acute crisis for Caracas.

The outcome may not be wholly negative for Beijing’s broader regional position. While China stands to lose a sympathetic government in Venezuela, the US move reinforces perceptions of American hegemony and unpredictability, potentially encouraging other Latin American governments to hedge by deepening ties with China.

Oil

Lots of the China-Venezuela coverage so far has focused on oil. Venezuela’s largest crude export destination has been China, and Chinese firms such as China National Petroleum Corp (中国石油天然气集团有限公司) have long been involved in Venezuelan extraction. After Washington tightened oil sanctions in 2019, China halted direct purchases of Venezuelan crude. The oil did not stop flowing to China altogether; instead, it was rerouted through independent traders via ship-to-ship transfers and often relabeled as Malaysian crude, allowing Chinese refiners to keep importing while giving Beijing plausible deniability.

But Venezuela’s oil importance to China should not be overstated. Venezuela accounts for roughly 4% of China’s crude imports, and the country’s overall economic weight is small relative to Beijing’s core energy interests in the Middle East and elsewhere. A US-approved government in Caracas could also plausibly make Venezuelan oil easier for China to access directly by removing the need for sanctions evasion altogether.

Share

Debt

The more consequential material concern for Beijing is probably debt. Venezuela is estimated to owe China roughly $13-15 billion. That exposure helps explain why, following the US capture of Venezuela’s president, China’s top financial regulator reportedly asked policy banks and major lenders to review and report their Venezuela-related risks.

The risk is not simply default, but reprioritization. As The Guardian noted, a government under heavy US pressure could choose to place American creditors and claimants ahead of Chinese ones, leaving Chinese banks to absorb losses. The situation is further complicated by opaque loan terms, oil-backed repayment structures, and the political leverage that often accompanies debt restructuring.

This is where a US-engineered political transition could become thorny for Beijing. Would a new, US-aligned government honor existing Chinese loans and contracts? Would Chinese firms retain access to assets and projects they financed? Or would they be squeezed out under the banner of political realignment? How these questions are resolved could directly affect the US-China relationship amidst its ongoing trade war.

Most coverage of China’s pandemic response has focused on its handling of COVID. Far less attention has been paid to what China has done in its aftermath, during which the country has been making interesting moves to prepare for the next large-scale biological threat.

Since 2023, Beijing has revised the Infectious Disease Law (IDL) and the Biosecurity Law and launched new frameworks like the Public Health Emergency Response Law (PHERL). Their rhetoric has also been increasingly telling, with criticism of the US’s pandemic response and self-proclamations of China as a global leader in pandemic oversight.

Pandemic prevention in China has moved from emergency reaction to long-term system design.

Chinese officials appear determined to ensure the next COVID doesn’t start within their borders. That determination increasingly stands in contrast to the United States, where public health institutional capacity has lost steam since 2020, especially during Trump 2.0.

Today’s installment examines governance initiatives, but this is only one part of a much larger ecosystem. Future pieces hope to explore PPE stockpiles, vaccine production, early-warning surveillance, research and lab standards, and the AI-bio crossover.

Main Takeaways

• The CCP looks to be taking pandemic risk seriously. After China’s public-health system was shown unfit for purpose when COVID hit, Beijing has now enacted some of the most actionable steps of any major country to bolster its pandemic-readiness system.
  

• COVID exposed how costly Beijing’s old instincts were: burying early signals, punishing whistleblowers, and relying on improvised crackdowns left the center blind and politically exposed. The new reforms try to fix this by giving local officials clearer rules, reporting guidelines, and more room to act early without fear of punishment. Beijing appears willing to trade some information-control for a more rule-bound, faster-moving system, though whether officials feel empowered to speak up remains uncertain.
  

• A more centralized domestic monitoring and command system gives China greater ability to manage potential outbreaks internally, reducing pressure to depend on international organizations. That avoids reputational costs and protects “face,” which helps explain why China can buy-in heavily to pandemic preparedness while still resisting meaningful collaboration or data sharing with groups like the WHO.
  

• Globally, Chinese state rhetoric casts the U.S. as the country that bungled COVID while downplaying its own early missteps. And Beijing is positioning itself as an international leader on health governance, especially for the Global South.

*Starting with “Recent Government Initiatives,” each section ends with a grade. Taken together, China earns a C+ overall, which is an improvement over the D I would have given it pre-COVID, though still shy of the B- I’d give the US.

Subscribe now

Roadmap of China’s Agencies

Since 2023, the major players in China’s pandemic readiness system have received new mandates, budgets, or planning documents to strengthen their roles.

At a high level, China’s system runs on a clear hierarchy. The State Council directs national strategy, the National Health Commission (NHC) leads implementation, and a network of technical and support agencies (at both federal and provincial levels) executes the work.

Key Players

State Council (国务院) — the top command centre in any outbreak. It activates the “Joint Prevention and Control Mechanism” (联防联控机制), created during COVID, to coordinate ministries across health, industry, and emergency management. Since 2023, the State Council has signalled an effort to bolster its coordinating role for pandemic response.

National Health Commission (国家卫健委, NHC) — China’s main health authority and a cabinet-level executive department of the State Council. It drafts and enforces key laws, oversees the China CDC and the National Health Emergency Response Center, and manages early-warning and emergency-medical systems.

National Administration of Disease Control and Prevention (国家疾控局, NADC) — created in 2021 to strengthen disease control and biosafety. It sets national standards for surveillance and builds modern early-warning/data systems. It’s one of the key additions of China’s post-COVID infrastructure.

China CDC（中国疾控中心） — the technical core of the system. It collects and analyzes infectious disease data, runs testing labs, and provides guidance to local CDCs. The CDC workforce numbers surged during COVID, increasing by about 20% to reach 240,000 in 2022, the highest level ever. This was preceded by years of post-SARS neglect, which left the system understaffed and unprepared for COVID (see graph below).

There are also many supporting ministries that handle logistics, funding, and research, such as

• National Health Emergency Response Center (国家卫生应急中心) - coordinates emergency medical teams and logistics during crises.

• National Biosecurity Work Coordination Mechanism (国家生物安全工作协调机制) - coordinates biosecurity-specific policy and emergency response across ministries.

• Ministry of Industry and Information Technology (工业和信息化部, MIIT) - manages medical supply production and logistics.

• Ministry of Science and Technology (科学技术部, MOST) - supports new R&D programs in pathogen detection and modelling.

• National Medical Products Administration (国家药品监督管理局, NMPA) - fast-tracks new countermeasures.

• People’s Liberation Army (中国人民解放军, PLA) - deploys medical units and runs military R&D in pandemic-related situations.

How Does the US Compare?

In China, authority flows from the State Council through the National Health Commission and its affiliated agencies. Provinces largely mirror this structure, which makes it easier to coordinate and implement national policy quickly once priorities are set in Beijing.

The US system is much less centralized. The Department of Health and Human Services — mainly through the CDC and the Administration for Strategic Preparedness and Response (ASPR) — leads at the federal level, but state and local governments hold most of the practical authority over public health measures. In practice, the federal government provides funding, guidance, and aggregates data, yet in a major pandemic, it’s less clear that the US could quickly coordinate a unified national response.

Share

Centralization, on the other hand, has trade-offs. China’s unified chain of command can move quickly, but a bad call at the top can misdirect the entire system. The U.S.’s decentralized model is more heterogeneous, since one state’s mistakes don’t necessarily drag everyone else down. China’s approach, therefore, relies heavily on accurate information flowing upward and on giving localities enough room to adapt policies to local conditions, which many of the initiatives below attempt to do.

Recent Government Initiatives

In September 2025, China’s top legislature (NPCSC seventeenth session) passed the Public Health Emergency Response Law (突发公共卫生事件应对法, PHERL). It’s the country’s most significant effort since COVID-19 to overhaul how it manages outbreaks, arriving alongside a substantial revision of the Infectious Disease Law (传染病防治法, IDL) earlier this year. Together, the two laws do a good job of weaving in many of the major pandemic readiness updates in recent years, and are meant to give China a more coordinated and legally coherent framework for handling future epidemics.1

One surprising feature of both PHERL and IDL is that neither substantively mentions the Biosecurity Law (生物安全法). Since its introduction in 2020/2021, the Biosecurity Law has been China’s main legal framework for managing biological risks, specifically, from pathogen labs to zoonotic disease surveillance. The law divides biotechnology research and development activities into three risk categories — high, medium, and low — requiring approval for high-risk and medium-risk activities. It also establishes classified management of pathogenic microorganisms and hierarchical administration of pathogenic microorganism laboratories. The law was mildly amended in 2024, though many of its weak spots remain.

The gist of these recent moves is an attempt to correct the legal and regulatory weaknesses that became apparent during COVID. At the time, SARS-CoV-2 was classified too slowly, lines of authority in emergencies were poorly defined, and rigid central control over information disclosure left local governments hesitant to act.

All Talk?

Do these reforms have any teeth? There are a few ways to parse this out.

The Biosecurity Law, IDL, and PHERL are binding laws (法律) passed by the NPC Standing Committee. This makes them more authoritative than the guiding opinions (指导意见) and plans (方案) that often crowd China’s policy space. Responsibility for their implementation also increasingly falls under the State Council, the top executive body of the land, giving these measures more political backing than if they were purely the responsibility of various lower-ranking ministries.

Still, the NPC passes many laws that aren’t effective. This is because (1) people don’t know they exist, or (2) they are not clear enough to be actionable. Therefore, what’s more important is enforcement clarity. Can local officials, hospitals, and labs actually understand what these laws require and act on them in real time? Here, the picture is mixed.

Many provisions are more explicit than previous drafts, but some remain vague or lack operational detail. For example, Article 74 of the IDL allows private entities to file complaints (申诉) if they believe emergency measures are excessive, a gesture to remediate the lack of voice many people felt during Zero-COVID. However, Article 74 offers little guidance on how such complaints will be handled or whether they provide meaningful recourse, making it unclear to people tempted to complain whether they will face consequences. By contrast, more fleshed-out stipulations like the updated early-reporting requirements (explained in the next section) clearly spell out responsibilities, timelines, and penalties, making them more obviously enforceable.

After reviewing the earlier versions of the IDL and Biosecurity Law and comparing them with the updated texts and the addition of PHERL, the system as a whole has gained some enforcement clarity. My rough sense is that only about 20–30% of the original provisions felt truly actionable, meaning that, as a local official or doctor, you could read them and understand what you were expected to do. In their current form, it feels closer to 40%.

Enforceability is jagged, though. The revised Biosecurity Law doesn’t feel meaningfully clearer to me, while PHERL and IDL seem to have made big strides.

Finally, we can look to historical analogues. The post-SARS reforms significantly reshaped China’s pandemic response system. Before 2003, the public health apparatus was fragmented and underfunded; the China CDC had only been established a year before, and case reports were still handwritten and faxed to Beijing. SARS prompted the government to carry out a wave of initiatives, such as building a real-time reporting network that linked clinics and hospitals across the country. The SARS reforms were incomplete, of course, given China’s lack of preparation for COVID, but it was a significant evolution from what little previously existed. The post-COVID reform wave feels like a similar energy stemming from a similar realization that their pandemic readiness system was far behind where it should have been.

The Content

*Graded from Best to Worst: Partly Post-COVID Improvements, Partly Overall Performance

Interagency Coordination

PHERL and IDL stress interagency coordination. The “joint prevention and control mechanism” (联防联控机制) created by the State Council during COVID is now written into law. It brings together more than 30 ministries and agencies across health, industry, and emergency management. At least a dozen civilian and military departments must share surveillance data, coordinate logistics, and build a unified national information platform for early warning. The aim is to keep ministries from working in silos and ensure outbreaks are met with synchronized mobilization.

Before COVID, no comparable command structure existed. Outbreak response rested with the NHC and China CDC, agencies without the authority to pull in heavyweight ministries or compel timely reporting from local governments. Coordination was improvised and slow. By placing the joint mechanism under the State Council, PHERL and the IDL give epidemic response a body that can enforce nationwide logistics and require all relevant ministries and provinces to report upward, ensuring that at least one institution has a complete, real-time picture of the entire situation.

Grade: A-

Mobilizing dozens of ministries and a national response is something the CCP can do better than anyone. The key caution is avoiding excessive uniformity; provincial conditions vary, and a highly centralized system must take care not to impose directives that could overlook unique situational circumstances.

Classification

The IDL updates China’s three-tier disease classification system (Classes A, B, C). Class A diseases, such as plague and cholera, trigger the highest-level emergency responses: immediate reporting, mandatory isolation, and broad quarantine powers. Class B diseases, like SARS or COVID (once it was officially listed), require strong but somewhat less sweeping interventions. Class C diseases, being the least concerning, are monitored primarily for trends and local containment, such as influenza or the mumps.

Previously, new or unknown pathogens couldn’t trigger a response until they were formally classified, a flaw made clear by how long it took to classify COVID. The revision tries to fix this by adding “sudden outbreaks of unknown origin [突发原因不明的传染病]” as an event that can be treated as Class A for response purposes. This designation prompts the State Council to rapidly investigate and issue a formal recommendation, allowing containment measures to begin before full classification is complete.

The concern for diseases of unknown origin reflects China’s growing rhetorical emphasis on “Disease X (X疾病)” (coined by the WHO in 2018), which calls for proactive preparation against future, as-yet-unidentified pathogens. As a government white paper put it earlier this year, China now aims to “draw on the experience of COVID-19 prevention and control, and make proactive preparations for future pandemics such as Disease X.”

Grade: A-

This lets officials act preemptively rather than reactively, but I’m docking half a grade as the incentives around sounding the alarm early are still uncertain. It’s unclear whether people will actually feel safe triggering a potential Class A response even when they’re technically allowed to do so.

Monitoring and Surveillance

Surveillance has taken on a more prominent role in the new framework. IDL Article 42 now mandates what’s called “sentinel surveillance” (哨点监测), a system in which selected hospitals and clinics continuously report data on specific diseases or symptoms to detect unusual spikes early. The revisions also strengthen requirements for identifying and reporting clusters of unknown or emerging illnesses, bringing China’s procedures more in line with the World Health Organization’s revised International Health Regulations (IHR).

Article 13 forbids excessive data collection and limits the use of personal information (like digital travel codes) to infectious-disease prevention and control. In theory, that’s a privacy safeguard; in practice, it’s anyone’s guess how strictly those boundaries will be enforced.

More speculatively, China’s ‘AI Plus’ Plan and related AI + Medical/Healthcare guidelines envision using artificial intelligence to enhance this surveillance network. The health-industry guideline lists public health services as one of four key application areas for AI, and pilot programs in cities like Shanghai are experimenting with AI systems that use citizens’ health data for lifelong health monitoring or proactive symptom detection. These efforts, however, remain largely aspirational.

Grade: B+

China already has the world’s most capable general surveillance system, so it will likely be able to implement this effectively. It’s still surprising that disease-specific surveillance measures weren’t firmly in place before COVID.

Local Authority

Under the IDL, local authority is also expanded. County- and city-level governments can now issue early warnings (Arts. 9, 53) and activate emergency responses when dealing with a sudden outbreak of unknown origin (Art. 65). This aligns the IDL with the Emergency Response Law, closing the gap between local initiative and national oversight. In theory, it allows quicker reaction on the ground while keeping reporting lines to Beijing intact.

Grade: B

Local officials can now move faster while Beijing deliberates — just not too fast, given an early move might look bad optically and provoke backlash from Beijing if it turns out to be a false alarm, given how vague the ostensible protections are.

Government Accountability

When it comes to checking central government power after some of the most controversial Zero-COVID measures — such as sealing residents in their homes, welding apartment doors shut, mass quarantine transfers, and imposing citywide lockdowns that lasted weeks — the recent reforms offer only modest adjustments. New provisions require local governments to ensure food and water supplies, maintain medical access, protect vulnerable groups, publish emergency hotlines, and keep workers employed during lockdowns (Arts. 64–67).

These steps are intended to prevent the worst excesses, but they do not meaningfully limit the state’s authority to impose sweeping restrictions in the first place. It is stated multiple times that decision-making remains centralized, and local officials must still carry out whatever directives Beijing issues.

Grade: B-

The CCP won’t be publicly apologizing for Zero-COVID anytime soon. But these reforms tacitly acknowledge its excesses and theoretically prevent future worst practices, like quarantined residents being locked in their homes without food.

Punishment

The revisions further strengthen enforcement but aim to channel it through clear legal authority. Individuals or institutions that refuse to cooperate with legitimate disease-control orders can now face fines of up to 1,000 yuan (~US$140), and entities up to 20,000 yuan (~US$2,810) (Art. 111 of IDL). Previously, Chinese law didn’t penalize most violations of epidemic orders, forcing police to repurpose unrelated statutes — such as those meant for constitutional “states of emergency” — to enforce zero-COVID restrictions. The fine is small, but “refusing to cooperate” is defined so broadly that even something like declining to wear a face mask could trigger a penalty.

Grade: C+

If this were aimed at punishing officials who bury crucial information — like those in Wuhan who hid early COVID signals — it would be a big upgrade. Instead, it mostly adds small fines that feel more suited to policing minor noncompliance, which risks echoing the punitive instincts of Zero-COVID rather than fixing the real failures.

Dual-Use Technologies

The Regulations on the Export Control of Dual-Use Items (中华人民共和国出口管制法), updated in late 2024, fold biological materials, technologies, and associated equipment into the same export-control framework that governs chemical, nuclear, and other sensitive goods. Under the new update, biological exports are managed through MOFCOM, under the State Council, which now appears to have greater authority over licensing and enforcement. Still, it’s unclear what exactly has changed — the specific list of what qualifies as “dual-use” biological items has yet to be clearly defined.2

Grade: C

This feels more about restricting what China sells abroad than about tightening its own safeguards around creating dual-use biological tools to begin with. It’s good as a nonproliferation measure, but the issue of creating clear research norms and controls over dual-use work inside China is still largely unaddressed.

Early Reporting

A core reform is early reporting. Under IDL, hospitals, blood banks, and local CDCs must report suspected outbreaks, clusters of unknown illness, or abnormal health events within two hours through the national Direct Reporting System. Those who report in good faith are protected from punishment (and eligible for some sort of award) even if their alerts later turn out to be wrong (Art. 51), while any official or institution that interferes with or delays reporting can now be penalized.

These provisions appear to respond directly to the early weeks of COVID-19, when local officials in Hubei delayed or suppressed information about the emerging virus — most infamously in the case of Dr. Li Wenliang, the Wuhan physician reprimanded by police for spreading “false information” after trying to warn colleagues about an unusual respiratory illness. Tragically, he later died from COVID.

However, it’s never really explained what disease reporting is supposed to include, and the promise of protection for “good-faith” reporting also feels fuzzy, since no one has defined what counts as good faith.

Grade: C-

If I were a doctor, I’d still be somewhat uneasy reporting early warnings. The protections are vague, and the precedent for punishment is much higher than in other countries.

Biotechnology Risks

The biggest shortcoming with China’s pandemic readiness system, in my opinion, is that it has not made substantial progress in addressing the safety risks posed by biotechnology — meaning the dangers that arise when genetic engineering, synthetic biology, or laboratory manipulation of organisms could unintentionally create or amplify biological threats.

The 2021 Biosecurity Law was the first statute that gestures at governance in this space. It formally divided biotechnology R&D into three risk tiers — high, medium, and low — with high- and medium-risk projects requiring approval or registration and restricted to legally incorporated domestic entities. The law also established security management rules for human genetic resources and biological resources.

The law was amended with updates that took effect in April 2024, but the changes appear largely procedural rather than substantive. There are still no specific ethical guidelines for biotechnology R&D; the three-tier risk system (high, medium, low) lacks concrete criteria for how projects should be classified; and the vague references to “relevant departments” (有关部门) leave unclear which agencies are responsible for what. In practice, this means ethical oversight is likely to devolve to institutional review boards or ministry-level discretion. These bodies vary widely in capacity, and because biotech research is competitive, institutions may have incentives to adopt more permissive review practices to maintain an edge.

This gap is likely related to the fact that Beijing also views biotechnology as a strategic growth sector. Much of the Biosecurity Law reads more like a biotech development agenda with biosecurity sprinkled on top. Article 5, for instance:

“The state shall encourage innovation in biotechnology, strengthen the building of biosecurity infrastructure and the biotechnology workforce, support the development of the bioindustry, raise the level of biotechnology through innovation, and enhance the capabilities to guarantee biosecurity.”

Grade: D

Synthetic pathogens are one of the most plausible routes to a truly catastrophic outbreak, yet Beijing’s biotech push largely ignores these safety concerns. However, in the US and other countries, ethical oversight also seems to fall to institutional review boards or ministry-level discretion, so I can’t give this a completely failing grade.

To sum up, China’s recent policy initiatives reflect a system trying to learn from its own COVID contradictions. Beijing wants a more unified and legally codified pandemic readiness system, one that detects and contains outbreaks before they spread, but also one that avoids repeating harsh crackdowns and which made provincial authorities feel powerless to national authorities. It’s a tough balance to strike.

Overall Grade: C+

Many of the laws and local incentives are still unclear, but Beijing is at least increasingly turning abstract goals into concrete procedures and has an unmatched capacity to trigger a unified response. And unlike some countries, its rhetoric does not actively denigrate public health measures.

For reference, I would give the US a B-. Even with its issues, the US does better at dual-use tech governance [pdf] and has better incentives for early reporting and information sharing.

Funding

Funding is an indicator of whether these statements have backing to them, but data is limited.

Our best piece of evidence comes from the Chinese Ministry of Finance, which shows that major infectious-disease prevention funding rose from about ¥16.98 billion ($2.38 billion) in 2018 to ¥23.82 billion ($3.34 billion) in 2023 — an increase of roughly 40% over five years. These funds are meant to expand things like vaccine-production capacity, surveillance systems, and hospital preparedness.

There’s no visible COVID-era spike in 2020–21 because much of that emergency spending flowed through temporary epidemic-response channels — one-off MOF transfers and provincial emergency budgets — rather than this regular subsidy line. The subsidies in the graph instead indicate Beijing’s effort to institutionalize emergency spending within its normal public-health budget.

Small bits of additional evidence tentatively point in the same direction. China launched a multi-billion-dollar reconstruction of the national CDC system in 2023, alongside major provincial investments in places like Shanghai and Guangdong. Data beyond 2023, however, is limited, so drawing further conclusions would be premature.

Grade: B-

They’re on an upward trend, but the total still looks modest relative to China’s GDP and population. In 2023, ¥23.8 billion (~US$3.3 billion) works out to only about US$2–3 per Chinese citizen per year. By contrast, OECD estimates put average pandemic prevention, preparedness, and response spending at around US$101 per capita, with the United States at US$279 and Germany at US$209. The true gap must be smaller, since China adds money through provincial budgets, immunization programs, and other health lines that don’t show up in the MOF subsidies, but I estimate that the sum of these monetary investments still falls well below the OECD average.

How China Talks About the US

One interesting factor shaping China’s pandemic governance has been its rhetorical positioning vis-à-vis the United States.

Beijing has leaned heavily on the narrative that America’s COVID response was chaotic, politicized, and unscientific, using US failings as a foil to validate its own system. The strategy deflects criticism of China’s early missteps and reinforces the idea that China’s centralized model is not only legitimate but superior.

For example:

• A People’s Daily editorial from May 2025 calls the US the “全球第一抗疫失败国” (literally: “world’s No. 1 failure in pandemic response”), citing CDC death totals and arguing that the outcome exposes pseudo-science.

• A post on the National Health Commission’s website accused the US of “squandering time” and policy ineffectiveness.

• A Global Times editorial said, “As the world’s most developed country, its response to the pandemic has been a complete failure, offering no positive lessons.”

This narrative has political uses, but it could also make Beijing overconfident. By defining itself in opposition to the US, China has built a pandemic story that depends on its own perceived success, which could also make addressing institutional shortcomings difficult.

Subscribe now

For instance, Chinese state media outlets have often disparaged the effectiveness of US vaccines, framing Western rollout efforts as reckless or unsafe. Yet beneath those critiques lies the unspoken acknowledgment that during COVID-19, China’s vaccine sector fell far behind its Western counterparts in both technology and trust. Beijing’s decision to reject mRNA vaccines like Moderna, despite their demonstrated efficacy, left millions reliant on weaker domestic shots.

Meta-Grade: China is grading its own paper here, and giving itself full marks despite doing a lousy job of handling COVID. Revising history, rather than addressing one’s mistakes, tends to be a bad idea.

International Moves

China has also engaged in a series of international initiatives on pandemic preparedness, though international communiqués on public health are rarely binding. Xi’s Global Security Initiative, for example, claims China will lead international biosecurity, but says little about how it will actually accomplish this.

Funding is a clearer signal. In May 2025, China pledged $500 million to the WHO over five years, effectively becoming the organization’s largest funder after the US withdrawal. China has also contributed to many other initiatives, like the World Bank’s Pandemic Fund, and hasn’t abstained from other multilateral health financing mechanisms (unlike the US).

A substantial portion of China’s health funding targets the Global South, particularly in Africa. China committed $80 million for constructing an Africa CDC headquarters in Ethiopia, a project that became operational during the pandemic, and $2 billion in assistance for COVID-19 response and economic recovery in developing countries. China’s WHO funding notably includes the condition of “a certain amount of voluntary contribution and projects support through the Global Development and South-South Cooperation Fund” — terminology tied to China’s Health Silk Road initiative, essentially the public health dimension of the BRI. 52 out of 54 African countries have participated in these health programs.

​​

China has also recently convened ASEAN Conferences on biosecurity governance in conjunction with the UN Office of Disarmament Affairs. These talks emphasize lab safety, pathogen-sharing, and early-warning systems between South East Asian countries.

Despite positioning itself as a global health leader, China consistently fails to report the specifics of its assistance activities to international agencies like the OECD’s Common Reporting System or the International Aid Transparency Initiative. Tensions have also flared with organizations like the WHO over China’s lack of timeliness, completeness, and durability of data sharing, especially around origins-relevant evidence for COVID and during recent disease surges. In November 2023, for instance, the WHO formally requested detailed information on pneumonia clusters among children following reports of cases in northern China. Beijing eventually provided data, but only after a significant delay, underscoring a pattern of reactive rather than proactive disclosure. China furthermore does not participate in Joint External Evaluation Assessments, where a team of independent international experts evaluates a country’s health security capabilities across 19 technical areas.

I believe this kind of behavior makes sense when accounting for how central reputation and “saving face” are to China’s public-health motivations. Reporting outbreaks quickly or exposing gaps in its own system can be embarrassing; projecting itself as a global public health advocate and generous benefactor to the Global South is not. If China can manage its own health problems internally and fund other systems externally, it (1) looks good and (2) reduces outside scrutiny — a bit like a boyfriend who pays for dinner so his girlfriend doesn’t go through his phone.

Grade: C

They say all the right things, and it’s good they’re helping the Global South’s public health infrastructure, but they still avoid building the deeper collaborative foundations we’d need for a globally unified response to a major infectious outbreak.

Share

Conclusion

The CCP is taking pandemic readiness seriously, but the through line isn’t a coherent strategy so much as a collection of post-COVID impulses: prevent another global pandemic from originating in China, avoid another round of draconian lockdowns, and do it all without loosening Beijing’s grip while empowering people to speak up.

Call to action

If you know anything about this topic or think I’ve missed something important, please reach out. I’m particularly interested in hearing from people with knowledge about China’s vaccine development capacity, high-end PPE manufacturing, biosurveillance systems, or research and safety standards for future installments.

I did not find nearly as many experts on Biosecurity x China as I would have liked. China’s pandemic preparedness apparatus remains surprisingly under-studied, especially compared to the extensive analysis of its COVID response or the pandemic readiness systems of other countries. The expert on this could be YOU.

Follow up to: nick@chinatalk.media

Nick Corvino joins ChinaTalk this year as a Tarbell Fellow, fresh off completing his master’s in China Studies at the Yenching Academy of Peking University.

Last week, China placed export controls on a wide range of rare earths and industrial inputs. Today, we’re taking a deep dive into the controls on lab-grown diamonds and why they matter.

Beijing’s restrictions target only industrial-grade synthetics used in chip fabrication and precision manufacturing. They can be important for wafer-slicing saws, polishing tools, and lithography optics, where extreme hardness and heat resistance are critical. Without them, producing advanced semiconductors and other high-tech components becomes significantly more difficult.

China is the world leader in synthetic diamond production, though by how much depends on whom you ask. Industry analyses suggest it accounts for roughly half of global output, while China Daily has claimed as high as 95% (this is too high..). It also dominates the manufacturing of key machinery such as high-pressure, high-temperature (HPHT) presses. This mix of capacity and machinery could let China squeeze parts of the chip supply chain, though its edge is less dramatic than some RREs like Terbium and Dysprosium.

What do diamonds do?

A diamond is made entirely of carbon atoms, each one tightly bonded to four others in a neat, repeating lattice. That structure gives it an uncommon mix of traits: it’s both the hardest material on Earth and one of the best conductors of heat. In practice, that means diamond can withstand enormous pressure without deforming and transfer heat faster than copper while remaining an electrical insulator.

These qualities are what brought synthetic diamonds into the semiconductor world. As chips for AI training grow hotter and denser, synthetic diamonds are becoming increasingly valuable for managing the resulting thermal load.

Since the 1950s, scientists have been able to make diamonds from scratch. These synthetic diamonds are molecularly identical to natural ones, but are grown in labs instead of forming underground. By recreating the extreme heat and pressure found inside the Earth’s mantle, or by building them atom by atom in controlled chambers, manufacturers can now produce crystals tailored for industry. That breakthrough allows the diamond industry to scale, no longer solely dependent on what can be dug out of the ground. Synthetic diamonds are now a critical component in slicing semiconductor wafers, printing chip designs, and improving radar capabilities (see the ‘Applications’ section for details).

Why China Leads in Synthetic Diamonds

China’s work on lab-grown diamonds began in the early 1960s, when researchers set out to make the country self-reliant in what were then called “superhard materials.” At the time, China couldn’t easily import industrial abrasives or natural diamonds, so it built its own high-pressure, high-temperature presses to produce them domestically. In 1963, China created its first synthetic diamond, becoming the fifth country in the world to do so. The effort echoed a broader post-war ambition to dominate foundational materials, such as rare earths and magnets, that technology and industry increasingly depended on.

Over the following decades, provinces like Henan and Shandong industrialized around diamond production, helped by cheap energy, access to carbon feedstocks (the raw materials that provide carbon atoms), and state support for superhard materials. By the 2000s, China was already producing the most synthetic diamonds in the world and had built an entire supply chain ecosystem around them.

China’s ‘Diamond Capital’

There are various places in China that have become the global hub for something highly specific. In the rare earth industry, the city of Baotou (包头市) in Inner Mongolia processes more than half of China’s rare earth minerals. Similarly, Dan Wang’s Breakneck describes how Zheng’an (正安) in Guizhou became “guitar city,” a small inland county that now makes about one in every seven guitars worldwide.

If Baotou is the rare-earth powerhouse, Zhecheng County (柘城县) in Henan Province plays the same role for synthetic diamonds. Home to fewer than a million people — very small, by Chinese standards — it has transformed from an agricultural county into China’s “Diamond Capital.” Local factories produce everything from micron-sized diamond powders to gemstone-quality crystals. Zhecheng now accounts for around half of China’s synthetic diamond output — roughly 4 million carats annually — and exports to more than 50 countries, amounting to an estimated 25–40% of global production.

Zhecheng also has the advantage of local clustering. Raw carbon feedstock suppliers, HPHT press makers, polishing workshops, and logistics firms all sit within literally a few square kilometers. That concentration lowers transaction costs, spurs reinvestment, and locks in process knowledge.

Overcapacity?

Just because China produces the most synthetic diamonds doesn’t mean it has the most leverage. In fact, the opposite may be true.

Like other parts of China’s manufacturing sector, the synthetic diamond industry could suffer from overcapacity, with too many producers chasing too few profitable markets, driving desperation to export abroad. Even the industry’s largest Chinese firms, including North Industries Group Red Arrow, Henan Huanghe Whirlwind, and Henan Liliang Diamond, reported sharp declines in revenue and profit in 2023, with their share prices staying relatively flat ever since and only modestly rising following the MOFCOM announcement. And Zhecheng County’s GDP per capita was ¥36,079 in 2024 (about $5,000), well below both the provincial and national averages. If China tightens export controls, it will inflict pain on domestic workers and manufacturers as well as overseas buyers.

How much do other countries depend on China?

How quickly could other countries scale production if China restricted exports?

Outside China, a handful of nations maintain smaller yet capable synthetic diamond industries focused on high-end or specialized applications. Element Six, a De Beers subsidiary headquartered in London, is the most prominent non-Chinese producer, manufacturing advanced diamond materials for cutting tools, optics, and semiconductor cooling. In East Asia, Japan has a few synthetic diamond companies, such as Sumitomo Electric and Tomei Diamond, and South Korea has ILJIN Diamond. India is taking a slightly different approach, favoring the slower, more expensive chemical vapor deposition process, which yields higher-purity stones than China’s majority HPHT method. Russia still leads in natural industrial diamonds — about 40% of global output — but that segment is increasingly obsolete compared to synthetic materials.

The United States has its own producers, including Hyperion Materials & Technologies and Applied Diamond Inc., but they lack China’s industrial scale. According to the U.S. Geological Survey, the U.S. remains heavily dependent on China for imports — roughly 77% of its industrial diamond supply — while another 8% comes from South Korea, 5% from the UK, and the rest from a long tail of countries.

Producing industrial diamonds at scale requires not only thousands of presses but also a dense network of coating, bonding, and classification facilities—and, crucially, the decades of process know-how now most concentrated in China. Western producers also face tougher environmental permitting for powder-handling and metal-bonding operations.

However, as diamond analyst Paul Zimnisky told ChinaTalk:

“In mainstream high-tech applications, the global industry is still in a nascent phase — so it is still to be determined who will ultimately be the biggest players in this space.”

China has not achieved a complete monopoly across the diamond supply chain, nor does there seem to be a segment or tool within the chain that only China can produce. Other countries can make key inputs, such as diamond presses, precision abrasives, and CVD materials, just not at the same scale. So, China’s leverage is real, but compared to entrenched sectors like rare earths, where it processes ~90% of global supply, the diamond industry has real global players who could likely step up to the plate and take advantage of the market opportunities export controls would open.

Applications

Diamond Saw!

The diamond wire saw (one of the diamond products China restricted) is a thin metal wire coated with fine diamond dust, used to slice huge blocks of silicon into the flat wafers on which chips are built. The process demands extreme precision. Each wafer must be perfectly smooth and even, because even microscopic irregularities can throw off the alignment of the billions of transistors that will later be etched onto its surface.

Not every wafer is cut this way. Older or lower-end manufacturing lines sometimes use slurry-based wire saws, where loose abrasives like silicon carbide are suspended in fluid to do the cutting. But over the past decade, diamond-coated wires have become the preferred choice, since they cut faster, waste less silicon, and leave cleaner, flatter surfaces. Synthetic diamonds, grown under controlled conditions, are uniform in size and hardness, allowing the wire to stay cool and cut with consistent precision.

That precision matters because every step of chipmaking builds on the wafer’s surface. Lithography machines must focus light with nanometer accuracy; deposition and etching layers have to align perfectly on top of one another. If the wafer is even slightly uneven or warped, this can mess up every subsequent step in the chip-making process.

(You can also buy a commercial-grade diamond saw at your local Home Depot!)

Optics and Lithography

Diamonds can also appear inside lithography systems, particularly in EUV machines that print the smallest and most advanced chips. The light inside these systems is so powerful that ordinary materials would warp or degrade, so manufacturers use diamond to handle the heat and maintain optical stability.

Diamonds can also be used to polish the mirrors and lenses inside lithography machines to near-atomic smoothness, a requirement for maintaining optical precision at nanometer scales, where even microscopic surface flaws can blur or misalign the projected chip pattern.

Wildcard: Diamonds as Future Chips?

Diamonds don’t just have to make semiconductors — they could one day become them. A semiconductor is a material whose ability to conduct electricity sits between a conductor and an insulator (hence the name ‘semi’ in ‘semiconductor’), allowing precise control over electrical current. Diamond’s wide band gap and exceptional thermal conductivity make it theoretically better than silicon for handling high voltages and extreme heat.

Researchers in Japan, the U.S., and China have already built prototype diamond transistors, but they’re still costly and hard to produce. Doping a diamond with boron atoms can turn it into a semiconductor, though the process remains difficult to control. More practically, companies like Diamond Foundry are experimenting with embedding small pieces of diamond into silicon chips to keep them cool and increase energy efficiency.

Military

Diamonds are also an important defense material beyond AI. Their hardness and thermal conductivity make them valuable for tooling munitions and cooling high-power radar and laser systems.

Unanswered Questions

After a week of research, here are the questions I couldn’t crack:

• How quickly could non-Chinese producers scale if restrictions bite? Other countries can do chemical vapor deposition and make diamond saws — but how long would it take to match China’s scale and cost advantage if they were suddenly asked to increase supply?

• Is there a single segment of the supply chain that only China (or another country) truly controls and can’t yet be replicated? My best guesses are:

  1. the human know-how and culture in Henan;

  2. HPHT press manufacturing and tooling (my current estimate is that China produces 80–90% of HPHT diamonds globally, though much of this output is low-cost and not high-enough quality for many industrial uses); and

  3. catalyst & ultra-pure graphite feedstocks, since China is really good at graphite refinement and metallurgy.

• Are governments planning for this? Who is implementing policies to reduce these vulnerabilities?

If you know the answer to any of these, or work in the diamond industry, I’d love to hear from you! Shoot me a message at: nick@chinatalk.media

Lily’s Notes on Diamond Jewelry

MOFCOM explicitly excluded jewelry-grade diamonds in the export controls. Why? It certainly would have hit Americans where it hurts — the USA is the world’s largest consumer of diamond jewelry, and in 2024, nearly half of all diamond engagement rings sold in the US contained lab-grown diamonds.

But Beijing is keen to argue that the new export controls are matters of national security, and not just an attempt to generate leverage in negotiations with the US. Restricting exports of gem-grade diamonds would undermine that argument, since they are very obviously not dual use.

Paul Zimnisky also noted that Indian producers are well-positioned to acquire market share in the wake of Chinese controls on industrial diamonds. If China also restricted exports of jewelry-grade diamonds, Indian suppliers would be incentivized to expand production to fill the gap. That would reduce the profitability of the nearly 1,000 Chinese firms that deal in lab diamonds — most of which are small businesses — even if the controls were temporary.

On diamond prices

In the wake of the new export controls, industrial diamond producers could pivot to producing gem-grade material (the playbook for this upgrade is not new). Such an abrupt increase in supply would decrease prices and eat into gem-grade diamond producers’ profit margins, as well as drive natural diamond prices even lower to compete.

But in fact, lab diamonds have already driven natural diamond prices down. In the words of diamond industry veteran Jon Phillips, “The diamond market has stabilized, but that doesn’t mean it’s not ripe for another downfall… [Natural diamonds] are inextricably tied to the lab-grown market.”

Chinese lab diamonds are available for 1/20th of the price of an equivalent natural diamond, with one-carat faceted stones available for as little as 1,000 RMB (~US$140). That price difference has helped lab diamonds grow from 1% market share in 2015 to about 20% today. That’s bad news for Botswana, which relies on diamond mining for 30% of its GDP, but certainly a win for enjoyers of sparkles the world over.

In addition to price and aesthetic considerations, there’s also an element of national pride that drives Chinese consumers in particular to choose lab diamonds. As Beijing-based jeweler Cheng Cheng 成诚 told ChinaTalk:

“From 2019 to 2023, many of my clients would ask about the difference between natural and lab-grown diamonds and tended to favor natural ones…. By 2024 and 2025, however, I’ve felt that many customers have become very comfortable with lab-grown diamonds. …

One reason for this shift is that lab-grown diamond manufacturing is centered in China. Confidence in “Made in China” products has grown rapidly, and Chinese consumers strongly recognize the country’s fast-developing lab-grown diamond technology.”

Some consumers will always prefer a gem that was formed over millions of years by geologic processes. But the beauty of lab-grown gemstones is that they expand the range of possibilities for artistic jewelry designs.