HomeEverythingEducation
Equities & FundsCrypto & Digital AssetsAI & TechnologyBusiness & CorporateUS Politics & PolicyGeopolitics & Global RiskMacro, Rates & FXCommodities & EnergyEuropean Politics & MarketsAsia-PacificReal Estate & Property
← All Stories

US firms turn to Chinese AI models amid US restrictions

Created at 6 Jul · 4:35 PM1 source↑ Market-relevant
IN SHORT

US companies increased their use of Chinese AI services in June following US government restrictions on access to top American models like Anthropic's Mythos. Open-source Chinese models offer significant cost savings and uninterrupted access, making them increasingly attractive to US corporations.

✉Newsletter

PiQ Daily

Pick your topics. Get only what matters, on your cadence.

Key Numbers

5%cost of Chinese AI models compared to Anthropic's Claude Mythos
one-twentiethcost of Chinese AI models compared to Anthropic's Claude Mythos
June 12date of US export controls on Anthropic models
July 1date of Fable model restoration

Who's Involved

US companies
increased usage of Chinese AI services
Anthropic
suspended top AI models due to US government restrictions
US Department of Commerce
imposed export controls on AI models
Emmanuel Macron
French President who criticized US AI policy
GLM
Chinese AI model provider
DeepSeek
Chinese AI model provider
Coinbase
US firm using Chinese AI models
Uber
US firm using Chinese AI models
US firms turn to Chinese AI models amid US restrictions

↳ Why This Matters

The U.S. government's restrictions on AI model access are inadvertently bolstering China's AI ecosystem by pushing international users, including American firms, towards Chinese alternatives. This shift could impact the global competitive landscape of AI development and adoption, potentially narrowing the lead of U.S. AI companies.

Key facts

  • US companies' adoption of Chinese AI services increased significantly in June.
  • US government export controls on Anthropic's Mythos and Fable models led to their suspension for non-US persons.
  • Open-source Chinese AI models are being adopted by US firms due to lower costs and guaranteed access.
  • French President Emmanuel Macron criticized the US action as "nationalist" and warned of consequences for US AI companies.
  • Mythos remains restricted to American institutions, while Fable has been restored with new restrictions.

U.S. companies significantly increased their use of Chinese artificial intelligence services in June, a trend driven by U.S. government restrictions on access to leading American AI models and the growing appeal of lower-cost, open-source alternatives from China.

The U.S. Department of Commerce on June 12 imposed export controls on Anthropic's high-end Mythos and Fable models, leading Anthropic to suspend access for non-U.S. persons. Unable to verify user citizenship, the company reportedly took the models offline for all users, including the U.S. National Security Agency.

This move marked a step backward for global access to advanced AI capabilities. French President Emmanuel Macron criticized the U.S. action as "nationalist," warning that such ad hoc controls could undermine trust in American AI companies and their global revenue streams. He noted that international clients are crucial for recouping the substantial costs of training frontier AI models.

While Fable was restored on July 1 with more restrictive guardrails, Mythos remains accessible only to American institutions. This episode, coupled with reports of OpenAI agreeing to U.S. government approval rights over customers for its latest model, highlights a shift towards more controlled access to cutting-edge AI.

The sudden U.S. action is expected to boost the adoption of Chinese AI models, which are often released as "open weight." This means users can run them on their own computing infrastructure, ensuring uninterrupted access once downloaded, a feature that has become a greater selling point following the U.S. restrictions. Chinese firms, however, offer no guarantee of future access and could adopt similar safety concerns.

Frequently asked questions

US companies increased their use of Chinese AI models due to US government restrictions on access to top American models and the lower cost and guaranteed access offered by open-source Chinese alternatives.

The US Department of Commerce placed export controls on Anthropic's Mythos and Fable AI models on June 12, leading to their suspension.

Chinese AI models can cost as little as 5% of the price of Anthropic's Claude Mythos, or about one-twentieth.

An 'open weight' model means users can run it on their own computing power, ensuring uninterrupted access once downloaded, unlike closed models where access can be revoked.

What Happens Next

01International access to Anthropic's Fable model was restored on July 1 with more restrictive guardrails.
02Anthropic's Mythos model is currently only available to American institutions.
03OpenAI has reportedly agreed to U.S. government approval rights over customers for its latest model.

Get the newsletter.

Pick the topics you actually care about. We'll email when there's news worth your time, on the cadence you choose. Cancel any time from your account.

Cadence

How It Developed

US companies' use of Chinese AI services surged in June.
Anthropic suspended some models due to US government restrictions.
Open-source Chinese AI models are becoming more entrenched in the US corporate world.
The US Department of Commerce placed export controls on Anthropic's Mythos and Fable models on June 12.
Anthropic took down the models for all users due to inability to verify citizenship.
French President Emmanuel Macron warned the US about "nationalist" actions impacting AI model access.
International access to Fable was restored on July 1 with stricter guardrails.
Mythos has been released only to American institutions for now.

Sources

T1
Chinese AI usage by US firms soared after Mythos restrictionsNikkei Asia
T2
Fable of the Mythos saga: Ad hoc US AI model controls could help China | PIIEpiie.com

Related Stories

Alibaba's Open-Source AI Gains Traction, Monetization Challenge Remains
6 Jul · 4:10 PM
Anthropic accused of secret user tracking in China
6 Jul · 4:50 PM
China's Biren seeks $900m for GPU development to rival Nvidia
6 Jul · 4:20 AM
Small businesses budget for AI's unexpected costs and errors
6 Jul · 9:25 AM
China's AI giants to remove chatbot personas amid regulatory crackdown
6 Jul · 5:45 AM