U.S. clears OpenAI’s GPT 5.6 as China weighs locking down its own AI models
The AI release race is colliding with geopolitics on two fronts today.
The AI release race is colliding with geopolitics on two fronts today.
In Washington, the Trump administration has lifted restrictions on OpenAI’s advanced GPT 5.6 model, clearing the way for a broad public release as soon as this week, Axios reports [1]. The decision follows extra testing by the Department of Commerce’s Center for AI Standards and Innovation and a series of meetings between OpenAI technical experts and government officials [1]. OpenAI had previously been forced into a staggered, government-approved rollout of GPT 5.6, a move the company said was not its preferred release strategy [1]. The episode mirrors earlier U.S. curbs on Anthropic’s Mythos and Fable models, underscoring how the world’s most powerful AI systems are now being released through case-by-case negotiations rather than clear, finalized rules [1].
Meanwhile, Beijing is moving in the opposite direction. Chinese authorities have held talks with Alibaba, ByteDance, and Z.ai about restricting foreign access to their most advanced AI models, including unreleased systems, according to a Reuters report cited by TIME [2]. Options on the table range from barring public release to limiting the models to domestic use only [2]. Such a shift would mark a sharp reversal for Chinese labs, which have gained global traction by releasing open-weight models for free [2]. The deliberations suggest China is now grappling with the same national-security trade-offs that led Washington to slow-roll GPT 5.6 and Anthropic’s models [2].
The parallel moves highlight a growing consensus among major powers: frontier AI models are too strategically sensitive to leave entirely to market forces.
Away from Silicon Valley and Shenzhen, the Middle East is flaring again. The U.S. military struck more than 80 Iranian targets, including over 60 Revolutionary Guard small boats, after Iran hit three tankers in the Strait of Hormuz, the Associated Press reports [3]. Iran retaliated with strikes targeting Bahrain and Kuwait, both home to major U.S. military installations [3]. The U.S. also revoked a license that had allowed Iran to sell oil openly on global markets as part of an interim ceasefire deal [3]. The attacks came during the multiday funeral for Iran’s late Supreme Leader Ayatollah Ali Khamenei, raising fears that already fragile negotiations for a permanent agreement could collapse [3].
Together, the day’s headlines show technology and hard power becoming inseparable: AI access is now a bargaining chip between governments, while a single chokepoint in the Persian Gulf can still rattle global energy security.