OpenAI will delay GPT-5.6 after Trump administration request
TL;DR
OpenAI is not launching GPT-5.6 broadly at first. The company is starting with a limited preview for a small group of enterprise partners, reportedly after a Trump administration request tied to security concerns. During the preview, the US government is reportedly involved in deciding which customers get access. OpenAI calls the arrangement temporary and says this kind of access process should not become the long-term default.
Nauti's Take
OpenAI is pitching GPT-5.6 as a performance leap, but the real story is access control. A short safety-focused preview can make sense, especially for cyber and bio capabilities.
The risk is that this turns into an informal licensing regime where proximity to government matters more than transparent rules. The industry needs hard criteria here, not PR-heavy language about safety and broad access.
Briefingshow
This is more than a routine safety review. If the US government effectively helps decide who can use a frontier model, AI release policy moves from internal safety governance toward state-mediated market access. Without clear standards, companies face uneven deals, slower launches, and political gray zones for developers, enterprises, and global partners.
Sources
