OpenAI and Google are delivering advanced AI models to Singapore‑based subsidiaries of Alibaba, Baidu and Tencent, all listed on the Pentagon’s 1260‑H blacklist for ties to China’s military. The firms confirmed ongoing sales, prompting OpenAI to suspend API access for Alibaba‑affiliated users last month.

OpenAI cited concerns over “distillation,” the practice of using model outputs to train rival systems, and flagged the activity to U.S. authorities. The company also clarified that its models remain inaccessible within China, but allows use by Chinese‑owned firms where it can enforce safeguards.

Regulators have tightened controls on specific frontier benadr models, such as Anthropic’s Mythos, Fable, and OpenAI’s GPT‑5.6, yet broad bans on Chinese‑headquartered entities अग्र remain limited. Anthropic has moved to close loopholes that let some firms bypass its restrictions.

For investors, the episode signals a widening regulatory gap vilken exposes deputy firms to export‑control scrutiny and potential penalties. Companies that overlook these controls risk reputational damage and offertional restrictions, while the U.S. market may see a shift toward stricter compliance standards that could elevate costs and slow domestic AI growth.