OpenAI and Google Reportedly Sold AI Models to Blacklisted Chinese Groups

OpenAI and Google have reportedly provided access to their advanced AI models to Singapore-based subsidiaries of Chinese technology companies that have been placed on the U.S. Department of Defense’s blacklist over alleged ties to China’s military. According to a Financial Times investigation, the transactions were conducted through overseas entities and appear to be legal under current U.S. export control rules, exposing potential loopholes in regulations governing the export of AI software. The report has intensified debate in Washington over whether restrictions on advanced AI models should be tightened alongside existing controls on AI chips.

The revelations come amid escalating U.S.-China competition in artificial intelligence, where governments are increasingly treating frontier AI models as strategic technologies with national security implications. While exports of advanced AI chips face strict controls, software access through overseas subsidiaries remains subject to different regulatory frameworks.

AI Services Reportedly Reached Blacklisted Chinese Groups

According to the Financial Times, Singapore-based subsidiaries linked to Chinese technology giants—including Alibaba, Baidu, and Tencent—were able to purchase or access AI services from OpenAI and Google.

Key findings include:

• AI services were provided to overseas subsidiaries rather than mainland Chinese entities.
• The parent companies are listed by the U.S. Department of Defense as firms allegedly linked to China’s military.
• The reported transactions complied with existing U.S. regulations.
• The arrangements have raised concerns about gaps in AI export controls.
• Policymakers are now examining whether current rules adequately cover AI software.

The Pentagon’s blacklist does not automatically prohibit all commercial dealings, unlike certain U.S. Commerce Department restrictions.

Why the Transactions Were Legal

Current U.S. export controls primarily target the export of advanced semiconductor hardware and certain restricted technologies.

According to the report:

• Overseas subsidiaries operating outside mainland China may still qualify for access under existing policies.
• AI software is regulated differently from advanced AI chips.
• Companies are generally expected to conduct compliance screening based on applicable sanctions and export-control rules.

This regulatory distinction has prompted calls for broader rules governing access to frontier AI models.

OpenAI and Google Respond

The report says OpenAI recently suspended access for some Alibaba-linked users after identifying concerns that its models could potentially be used for AI model distillation, a technique in which outputs from one model are used to improve another model.

Google reportedly maintains similar policies restricting access in certain jurisdictions but acknowledged that enforcement can be challenging when customers operate through overseas entities or subsidiaries.

Debate Over AI Export Controls Intensifies

The disclosures have renewed debate over whether advanced AI software should face restrictions similar to those already imposed on high-end AI chips.

Supporters of tighter controls argue that:

• Frontier AI models have strategic national security value.
• Software can be as important as hardware in AI development.
• Existing rules leave potential regulatory loopholes.
• Stronger oversight could help prevent unintended technology transfers.

Others caution that overly broad restrictions could affect legitimate multinational business operations and global AI collaboration.

AI Becomes a Geopolitical Battleground

The report comes as both the United States and China continue strengthening controls over advanced AI technologies.

Recent developments include:

• The U.S. expanding restrictions on advanced AI chips and related technologies.
• China considering tighter controls on overseas access to some of its own frontier AI models.
• Growing concern from both governments over AI’s military and strategic importance.

These parallel efforts illustrate how AI has become a central focus of technology and national security policy.

What Investors and the Industry Will Watch

Market participants will closely monitor:

• Possible changes to U.S. AI export regulations.
• Additional compliance requirements for AI providers.
• The impact on multinational technology companies.
• Future licensing rules for frontier AI models.
• Regulatory responses from China and other countries.
• Enterprise demand for cross-border AI services.

Any tightening of export controls could reshape how global AI companies provide access to advanced models.

Outlook

The report that OpenAI and Google supplied AI services to overseas subsidiaries of Chinese companies on the Pentagon’s blacklist highlights the growing complexity of regulating artificial intelligence in a global market. While the reported transactions appear to have complied with current U.S. rules, they have exposed potential gaps between controls on AI hardware and AI software.

As competition between the United States and China over AI leadership intensifies, governments are expected to face increasing pressure to update export-control frameworks. Future policy changes could significantly affect how frontier AI models are distributed internationally, balancing national security concerns with commercial innovation and global technology collaboration.