US artificial intelligence companies OpenAI and Google have been supplying advanced AI services to Singapore-based subsidiaries of Chinese technology giants that have been placed on a Pentagon blacklist, raising fresh questions in Washington over whether existing export controls are sufficient to slow China’s AI development.
According to a report by the Financial Times, the companies confirmed that they have provided AI services to subsidiaries linked to Alibaba, Baidu and Tencent, firms that the US government has accused of having ties to China’s military. The sales are legal under current rules, but the disclosures have renewed calls for stricter controls on access to advanced AI models.
The Financial Times reported that OpenAI suspended access last month for Alibaba-affiliated users of its API, the software interface that allows developers to remotely access AI models. The company said the move followed concerns over possible “distillation,” a process in which developers use outputs from one AI model to improve another system.
OpenAI said the suspected activity had been reported to the US government. The company said it does not allow its models to be accessed directly in China but permits some companies with Chinese ownership or headquarters to use its tools in countries where safeguards can be enforced and potential misuse monitored.
“We would rather see more of the world using AI shaped by democratic values than AI controlled by autocratic governments,” OpenAI said, while adding that nationality alone should not determine access to its technology.
Google said its AI services are available in Hong Kong and Singapore, subject to its usage policies, including restrictions against distillation. However, the company said geographic restrictions alone were not enough to address the risk because sophisticated actors could potentially bypass such limitations.
The issue has drawn attention because the US government has taken steps to restrict access to certain advanced AI models but has not broadly prohibited Chinese-headquartered companies, including those on the Pentagon’s so-called 1260H list, from using advanced AI software through overseas subsidiaries.
“The administration says we need to beat China on AI all the time, but the problem is they haven’t done anything on export control, which is the actual tool we have to slow China down,” said Chris McGuire, a technology and security expert at the Council on Foreign Relations, according to the Financial Times.
Alibaba, Baidu and Tencent did not provide immediate comments on their use of US AI services. Alibaba has separately challenged its inclusion on the 1260H blacklist, asking a US court to remove the designation and describing it as “arbitrary and capricious.”
While OpenAI and Google have allowed some overseas access under safeguards, AI company Anthropic has taken a stricter approach by banning Chinese companies and foreign entities owned by them from using its advanced models. Anthropic has also moved to close loopholes that it says allowed some Chinese companies to bypass restrictions.
Anthropic has previously accused Chinese AI firms, including DeepSeek, Moonshot and MiniMax, of using distillation methods. The company also warned Congress that Alibaba had allegedly used thousands of accounts to generate millions of exchanges with its Claude AI model in violation of its terms of service. Alibaba did not respond to the allegations at the time.
AI policy specialists have argued that stronger export controls are needed to prevent Chinese laboratories from gaining access to advanced AI capabilities without bearing the same development costs as US companies. Joe Khawam, a specialist in AI policy and national security law at the Law Reform Institute, warned that continued access could weaken the economic foundation of US leadership in frontier AI development.
