Breaking
Trump Administration Weighs Shaming Companies That Use Chinese Open-Weight AI: What We Know
Reuters reported Wednesday, citing three people, that Treasury and the White House science office have discussed making US companies post on their websites when they use Chinese AI models. It is a discussion, not a rule. Reuters described it as a “name and shame” approach within a push against Chinese open-weight models. Our analysis, labeled as opinion below: it is government pressure on open-weight and self-hosted AI, and it steers businesses toward a few closed US labs. Jump to our take.
The short answer
- What was reported: Reuters reported on Nov. 18, citing three people, that the Treasury Department and the White House Office of Science and Technology Policy (OSTP) have discussed requiring US companies to publicly disclose their use of Chinese AI models. Under the plan, companies like Airbnb would post that information on their websites.
- What it is: a “name and shame” approach, in Reuters' words: discourage reliance on Chinese models without new restrictions that could anger industry or Beijing.
- Status: a discussion, not a rule. Reuters said it was not clear the idea is gaining traction or would go through formal rulemaking. The White House, Treasury and Alibaba did not respond to Reuters.
- Our angle (opinion, see Our take): we read this as government pressure on open-weight and self-hosted AI, and a nudge toward a few closed US labs. That is our analysis, not Reuters' reporting.
- Named example: Airbnb, which uses Alibaba's Qwen for its AI customer service agent. Airbnb told Reuters its AI “runs overwhelmingly on U.S.-origin models.”
What we know
- The report. Reuters' Alexandra Alper reported the story as an exclusive on Nov. 18, based on three people. Reuters' own page was behind a sign-in when we checked, so we read the full text in syndicated copies from Yahoo, The Globe and Mail and Bloomberg Law.
- Who discussed it. The Treasury Department and OSTP, per Reuters' sources. Reuters said the plan “would apply to companies such as Airbnb.” It did not report who else would be covered.
- The mechanism. Disclosure on company websites, per Reuters. It would stop short of a ban or new restrictions. Reuters said it fits a broader push against Chinese open-weight models, a “source of deep concern within the government,” while the administration keeps its anti-regulation stance on AI.
- The models in question. Reuters named Moonshot AI's Kimi, DeepSeek and Alibaba's Qwen as examples that China hawks fear will “hook” US companies into dependence. Open-weight models make their core components public for anyone to download and adapt.
- Airbnb's response. In a statement to Reuters, Airbnb said it uses “only a limited number of China-origin models, all of which are open-source and run only through approved U.S.-based service providers, keeping data and operations separate and protected.”
- Beijing's response. The Chinese Embassy in Washington told Reuters that Beijing opposes politicizing trade and tech issues, which it said will “only stifle global AI advances and serve no one's interests.”
- Reaction from Congress. Rep. John Moolenaar (R-Mich.), chair of the House Select Committee on China, welcomed the deliberations. His committee posted a longer version of his statement on X: “Disclosure is an important step in managing the potential risks posed by the diffusion of Chinese open-weight models in the United States,” he said, adding that risks remain if users run the models on their own infrastructure. The statement also says Chinese models “may have embedded security risks the CCP can exploit.” That is his claim; read the post.
- The House investigation. Moolenaar and House Homeland Security Chairman Andrew Garbarino (R-N.Y.) opened a joint investigation on April 29 with letters to Anysphere (maker of Cursor) and Airbnb. They wrote to DoorDash on July 31. Reuters says they launched the probe this year; the letters are public: Airbnb, Anysphere, DoorDash.
- Where officials have stopped. Reuters wrote that US officials “have stopped short of restricting or taking punitive action against Chinese AI companies.” Treasury Secretary Scott Bessent said in July the administration could sanction Chinese AI companies found guilty of intellectual property theft through distillation, Reuters noted. It added that OSTP Director Michael Kratsios made similar accusations in April.
- Usage. Reuters said Chinese AI models were the three most popular by token usage in October on OpenRouter. OpenRouter counts only traffic routed through its own platform, not vendors' own APIs, cloud platforms or on-premises use, as Yicai notes, so it is a gauge of developer demand, not total market share.
- A government precedent. Reuters reported in September that the Federal Register website, run by the National Archives, offered a search tool using Alibaba's Qwen, even though the FBI had accused Alibaba of “malicious” copying of Anthropic's technology. The tool was taken down after social-media posts noticed it. Experts told Reuters the small model posed no substantial risk, according to Ars Technica.
What we don't know
- Whether it will happen. Reuters said it was not immediately clear if the idea was gaining traction inside the administration.
- Legal form and authority. Whether it would be a formal rule, guidance, an executive action or something else was not reported. Neither was which agency would write or enforce it, or what penalties, if any, would apply.
- Scope. Airbnb was the only company named as an example. Reuters did not report size thresholds, which industries, or whether internal uses (such as coding tools) count as well as customer-facing ones.
- What counts as a “Chinese model.” The report did not define it. Kimi, DeepSeek and Qwen were named as examples. Whether models built on top of them, such as a fine-tune, would count is not stated.
- Self-hosted versus hosted. Airbnb says its China-origin models run through approved US-based providers. Moolenaar says risks remain even when a company runs the models itself. How a disclosure rule would treat either setup was not reported.
- Timing. No date, draft or deadline has been reported. The White House and Treasury had not commented to Reuters.
Who is named
- Airbnb. Named by Reuters as an example. Its AI customer service agent uses Alibaba's Qwen. The House committees' April letter said Airbnb had stated publicly that it relies on Qwen over American alternatives because it is “fast and cheap.” Airbnb's statement to Reuters is above. (letter)
- Anysphere (Cursor). Reuters said Cursor's Composer 2 model was built on a version of Kimi, and that SpaceX bought Anysphere. The House letter calls Composer 2 “reportedly built on an open-weight model developed by Moonshot AI.” Anysphere and SpaceX did not immediately respond to Reuters. (press release)
- DoorDash. Reuters said a co-founder said the company routes some tasks in an internal system through Kimi. The House letter says DoorDash's Andy Fang said on July 6 that lower-level code-review work went to Kimi K2.6. DoorDash did not immediately respond to Reuters. (letter)
- Not reported as covered. Reuters named only Airbnb as a company that the plan would apply to. Cursor and DoorDash appear in the report as targets of the House investigation, not of the disclosure plan.
Why companies use these models: price
Reuters said Chinese open-weight models are gaining ground with US companies because they are cheaper than proprietary tools from OpenAI and Anthropic and “perform well enough.” Here is what a typical job costs at list prices, using the same 50,000-tokens-in, 5,000-tokens-out task on our API pricing table (checked Nov. 18):
| Model | List price per 1M tokens (in / out) | 50K in + 5K out |
|---|---|---|
| Claude Fable 5.2 | $8 / $40 | $0.60 |
| GPT-6.2 Astra | $8 / $40 | $0.60 |
| Qwen4-Max (Alibaba Cloud) | $1.50 / $4.50 | $0.098 |
| DeepSeek V4 Pro (peak-hour rate) | $1.32 / $3.96 | $0.086 |
Illustrative, not what any named company pays: list prices for standard prompts under 200K tokens, from Anthropic, OpenAI, Alibaba Cloud (Singapore) and DeepSeek. DeepSeek's off-peak rate is half. Cost = 50,000 × input price + 5,000 × output price. The models that led OpenRouter's usage ranking in October were other, cheaper Chinese models. Qwen4-Max is a closed model sold only through Alibaba's API (no weights published, per The Next Web), so it is not an open-weight model; DeepSeek V4 Pro's weights are open. Reuters' report concerns Chinese models generally, with open-weight models the stated worry.
At those prices the task costs about six to seven times less on the two Chinese models. Whether a disclosure requirement would change that math is an open question; Reuters reported no estimate of its effect. Our open-weights tokens-versus-spend report has the usage and spending data on open models more broadly.
The government's case
The government's stated worries are not frivolous. Rep. Moolenaar says Chinese models “may have embedded security risks the CCP can exploit.” In September 2025 the Commerce Department's CAISI evaluated three DeepSeek models and reported that agents built on its most secure DeepSeek model (R1-0528) were 12 times more likely than evaluated US frontier models to follow malicious instructions, that the same model responded to 94% of overtly malicious requests under a common jailbreak technique (versus 8% for US reference models), and that DeepSeek models echoed four times as many misleading CCP narratives as US ones. Those were 2025 models, not today's. Reuters also reported that OSTP Director Michael Kratsios accused China in April of stealing US labs' intellectual property on an “industrial scale,” that Treasury Secretary Scott Bessent said in July the administration could sanction Chinese AI companies found guilty of IP theft through distillation, and that China hawks fear Chinese models will make US companies dependent on Beijing. Reuters also listed American open-weight options: Meta's Muse Glimmer, Nvidia's Nemotron and Reflection AI's Beam.
Context: what Washington has already tried on open weights
- July 2024: NTIA says don't restrict open weights. The Commerce Department's NTIA recommended that the government not restrict the wide availability of open model weights “at this time,” while monitoring risks and keeping the option to act later. It said open models “decentralize AI market control from a few large AI developers” and let users run models “without sharing data with third parties.”
- January 2025: the AI Diffusion Rule leaves open weights alone. The Biden-era Commerce rule controlled certain closed model weights but said it was not imposing controls on open-weight models, because their benefits outweighed their risks “for now.” Commerce rescinded the rule in May 2025.
- July 2025: the AI Action Plan backs open models. The Trump administration's AI Action Plan has a section titled “Encourage Open-Source and Open-Weight AI.” It says the federal government “should create a supportive environment for open models” and that “We need to ensure America has leading open models founded on American values.”
- 2025: Hawley's decoupling bill. Sen. Josh Hawley (R-Mo.) introduced S. 321 in January 2025. It would ban importing AI technology or intellectual property developed in China, with criminal penalties. The bill does not mention model weights; The Register reported that it could reach downloading Chinese models, and that how it would be enforced on freely available open models was unclear. It was referred to the Judiciary Committee and has seen no further action, per Congress.gov.
- October 2026: the China FIREWALL Act. Reps. Josh Gottheimer (D-N.J.) and Nick LaLota introduced H.R. 10670 on Oct. 1. It would bar federal agencies from using or buying AI applications from a “foreign entity of concern,” including software built on its foundation models. It covers federal systems and procurement, not private use, and has had no committee vote.
- July 2026: harder options reportedly floated. Axios reported, citing unnamed sources, that parts of the administration had tried “de facto bans on foreign open-source models” before, and that options considered included adding Chinese AI labs to the Commerce Entity List and an NSA and National Cyber Director advisory (both last year), an executive order on hosting Chinese models, and draft Commerce rules (last summer). Axios said officials wary of regulation killed all of them, but that national-security hawks had grown louder. Neither the White House nor Commerce responded to Axios. We could not independently confirm these accounts.
- July 2026: industry pushback. TechCrunch reported that Hugging Face, Meta, Microsoft, Mistral and Nvidia, among others, signed an open letter urging policymakers not to impose “premature restrictions” on open-weight models. TechCrunch noted that OpenAI, Anthropic, Google DeepMind and SpaceX were not among the signers, and that the signers have their own economic stake in open models.
Our take: this is pressure on open-weight AI
Opinion and analysis. Everything above this box is reporting. Everything below it is Underclassic's argument, not Reuters', and not established fact. Where we rely on someone else's claim, we say whose.
Our read: this is the government leaning on companies to stop running open models, with a transparency label as the lever. Reuters itself calls the plan a “name and shame” approach meant to “discourage reliance” on Chinese models, and places it inside a push to combat the spread of Chinese open-weight models. There is no law, no vote and no text. Reuters said it was not clear whether it would even go through formal rulemaking. A label never has to ban anything to work. If a company expects a public warning on its website, and maybe a letter from a House committee, for using Qwen, Kimi or DeepSeek, its lawyers can move it to a closed API long before any court tests the policy. One source told Axios that what is actually happening is “slower and more durable” than a ban, citing procurement rules, Entity List threats and public pressure campaigns aimed at US companies using Chinese models. Another source said such approaches would have “a chilling effect” on Chinese open-source tech and the US companies that rely on it. A name-and-shame label fits that pattern.
Self-hosting is in the crosshairs. Airbnb's defense is the standard one: China-origin open-source models, run only through approved US-based providers, with data kept separate. Moolenaar told Reuters that risks remain even if US companies run Chinese models on their own infrastructure. We think that sentence matters. If downloading weights and running them on your own machines, with no data leaving the building, still counts as a risk, then the objection is not about where your data goes. It is about where the model came from, and about who is allowed to run which software. Reuters did not report whether a disclosure rule would reach self-hosters. But Moolenaar's statement tells you which way the concern runs.
Why open weights matter. They are the one kind of AI you can download, run offline and keep on your own hardware, with no vendor in the loop and no prompt leaving your network. Washington has said as much. NTIA said open models cut the need to share data with third parties and spread market control beyond a few large developers. The AI Action Plan said many businesses and governments “have sensitive data that they cannot send to closed model vendors.” Open weights are also testable by outsiders, as CAISI's DeepSeek evaluation shows. That does not make them safe. You cannot read a model's behavior off its weights the way you read source code, and the CAISI results show real weaknesses in the DeepSeek models it tested. But only open weights let anyone outside the lab check.
Centralization has a price tag. On our table above, the same 50K-in, 5K-out job costs about $0.60 on Claude Fable 5.2 or GPT-6.2 Astra and about $0.086 on DeepSeek V4 Pro, whose weights are open: roughly seven times more. Qwen4-Max, at $0.098, is closed, so the cleanest open-weight comparison is DeepSeek. Cheaper closed tiers exist: OpenAI's GPT-6.2 Sol, at $2 and $10 per million tokens in and out, would run $0.15, about 1.7 times DeepSeek's cost. These are list prices for API access; self-hosting adds hardware costs of its own. See the full API pricing table. Nudging companies off these models does not send them to American open ones by default. Axios reported that US alternatives are “still limited at best” compared with cheaper, more practical Chinese models. For many businesses the practical alternative is a closed API from a handful of labs such as OpenAI, Anthropic and Google.
Who benefits. We have not seen evidence in any report that OpenAI or Anthropic lobbied for this proposal, and we saw no comment from either on it. What is on the record: one source told Axios that leading AI labs or their allies approach the administration every three to five months with ideas to ban open-source models, and Axios quoted David Sacks, an outside White House AI adviser, writing on X in July that “the leading closed labs, already a duopoly in terms of AI model revenue, want the government to eliminate their open-source competition.” Those are claims by an anonymous source and by Sacks, not findings. TechCrunch reported that OpenAI and Anthropic have urged the administration to respond to alleged IP theft by Chinese labs, and that neither signed the open-weight letter. Companies that did sign, such as Nvidia and Microsoft, also have commercial interests, as TechCrunch noted. Whoever pushes for it, the effect would be the same: fewer practical alternatives to a few closed labs, whose prices are several times higher.
The inconsistency. The same administration's Action Plan told the government to support open models. It also said they should be “founded on American values.” Axios reports that US companies are turning to Chinese open models because they are cheaper and “just about as good.” Backing open weights in principle while shaming the companies that use the cheapest ones is hard to square.
What we would want instead. If a specific model is unsafe, publish the testing, as CAISI did, and let companies decide with the evidence in hand. Keep security requirements for federal systems, where bills like H.R. 10670 are already pending. Do not treat a download as a confession. The question we keep returning to: should running open weights on your own machine ever need a label?
The backdrop
Reuters noted the consideration comes as revelations about Nvidia CEO Jensen Huang's affair with an alleged Chinese spy have fueled renewed calls in Washington for a tougher stance on China. We covered that story in what we know about Huang's resignation. The report also came out on X fast: @MTSlive's post summarizing it had more than 300,000 views by Wednesday afternoon.
Get the daily issue
One email a day. Unsubscribe anytime.
An Underclassic breaking story, updated as reporting develops. Daily briefing: underclassic.com. Related: AI API pricing per 1M tokens; open-source AI guide; open weights: tokens vs. spend. Sources: Reuters, Nov. 18, 2026 (read via Yahoo, The Globe and Mail, Bloomberg Law); House Select Committee on China on X; House Homeland Security Committee, April 29, 2026; July 31, 2026; letters to Airbnb, Anysphere and DoorDash; Reuters, Sept. 17, 2026; Ars Technica; OpenRouter rankings; Yicai Global; The Next Web on Qwen4-Max; NTIA open-weights report, July 2024; AI Diffusion rule; AI Action Plan; NIST CAISI on DeepSeek; S. 321; H.R. 10670; Axios, July 20, 2026; TechCrunch, July 24, 2026. Reuters' own page was not readable by our tools; every Reuters claim above is attributed and was checked against the syndicated full text.