OpenAI and Google Quietly Sell AI Tools To Chinese Firms

2026-07-10 · Source: AutoGPT · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation, AI Governance & Policy · Depth: Fundamental Awareness, quick

Summary

OpenAI and Google have confirmed they supplied advanced AI services to Singapore-based units of Alibaba, Baidu, and Tencent, all of which are on a Pentagon watch list. This practice, reported by the Financial Times on July 10, 2026, is legal under current U.S. rules, which permit access outside mainland China. OpenAI monitors usage, suspending Alibaba's API access last month for "distillation" and reporting it to the U.S. government. Google also allows access in markets like Singapore, enforcing usage rules against distillation, though acknowledging geographical limits are not foolproof. Anthropic, however, bans Chinese firms entirely and advocates for stricter export controls on AI software, similar to those for computer chips.

Key takeaway

For policy makers reviewing U.S. AI export controls, this situation highlights a critical loophole where Chinese firms on a Pentagon watch list can legally access advanced AI tools via overseas subsidiaries. You should consider expanding current restrictions to cover AI software more comprehensively, mirroring existing controls on advanced computer chips, to prevent potential technology transfer risks and ensure national security objectives are met.

Key insights

U.S. AI export rules contain a loophole allowing Chinese firms to access advanced tools via overseas subsidiaries.

Principles

• U.S. AI export rules have geographical loopholes.
• AI model distillation poses a significant threat.
• Stricter AI export controls are being advocated.

Method

OpenAI monitors API usage for signs of distillation, suspending access and reporting suspicious activity to the U.S. government.

In practice

• Utilize overseas subsidiaries for legal AI tool access.
• Implement strict usage policies against model distillation.
• Advocate for expanded AI software export controls.

Topics

• AI Export Controls
• Geopolitics
• Model Distillation
• Chinese Tech Firms
• OpenAI
• Google
• Anthropic

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Tech Journalist, Legal Professional, Policy Maker

Related on AIssential

• AI Weekly Issue #506: Washington Blocked One AI Lab. China Blacklisted 56 Companies. — AI export controls are now a mutual weapon, impacting global technology and supply chains.
• Deepseek ban? — China is weighing an AI export ban, potentially shifting from global market penetration to domestic isolation in response to US controls.
• Kimi K3 is no reason for China panic — China's open-source AI strategy for non-"frontier" models is a geopolitical play, likely to shift as capabilities advance.
• US to restrict American companies’ use of Chinese AI models — The US plans to restrict American companies' use of Chinese AI models due to national security, while China considers reciprocal restrictions.
• Encryption, spyware, and now Mythos: History shows why cyber export control doesn’t work — Export controls on dual-use cyber technologies, including advanced AI models, have historically proven ineffective due to circumvention and global proliferation.
• US ends ban on exports of Anthropic’s advanced AI — Advanced AI capabilities, particularly those with dual-use potential, are subject to evolving national security concerns and regulatory uncertainty.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AutoGPT.