The advanced AI models of OpenAI and Alphabet Inc.‘s Google (NASDAQ:GOOGL) (NASDAQ:GOOG) are reportedly being sold to Chinese tech giants previously blacklisted by the Pentagon.
Google and OpenAI have confirmed their provision of AI services to Singapore-based subsidiaries of Alibaba Group Holding Ltd (NYSE:BABA), Baidu Inc. (NASDAQ:BIDU), and Tencent Holdings (OTC:TCEHY), according to the Financial Times on Friday. These Chinese firms have previously been accused by the U.S. government of having ties with China’s military.
OpenAI told the publication that it suspended API access for Alibaba-affiliated users last month over suspected misuse. The company reiterated that its models are not available in China, though some Chinese-owned firms can access them in countries where OpenAI can enforce safeguards and monitor for model ‘distillation.’
Meanwhile, Google said that its AI services are available in Hong Kong and Singapore under policies that prohibit distillation, but acknowledged that geographic restrictions alone cannot prevent sophisticated attackers from bypassing them.
OpenAI and Google did not immediately respond to Benzinga’s request for comments.
US-China AI Tensions Deepen
The sales, while legal, have renewed calls for stricter U.S. regulation of advanced AI models. Although the U.S. has restricted access to some cutting-edge models, it has not broadly banned Chinese companies, including those with alleged military ties, from using advanced AI software.
However, in June, the Pentagon expanded its blacklist to include companies suspected of having ties to China’s military or defense-industrial sector, including Alibaba, Baidu, and others. This move barred the Department from contracting directly with these companies and procuring their products or services through third parties starting in June 2027.
At the same time, China has been considering imposing restrictions on the most advanced AI models, including both closed-source and open-source systems, to tighten its grip on advanced AI. Chinese authorities have reportedly discussed with Alibaba, ByteDance, and Z.ai the possibility of restricting overseas access to the country’s top AI models, including unreleased versions.
The development follows reports that Alibaba has banned employees from using Anthropic’s AI tools over security concerns, replacing them with its in-house AI assistant, Qoder. It also comes as the Trump administration raises national security concerns about advanced AI being misused by foreign military and intelligence agencies.
Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors.
Image via Shutterstock
Login to comment