OpenAI Restricts AI Model Rollout at US Government's Request
Limited preview of GPT 5.6 series amid national security concerns, shifting from previous direct public releases
Quick Look
- OpenAI has limited the rollout of its GPT 5.6 series (Sol, Terra, Luna) at the US government's request, citing national security.
- This marks a shift towards tighter AI oversight by the Trump administration, contrasting with its earlier minimal regulation stance.
AI-generated summary
Why It Matters
The US has recently increased AI oversight citing security concerns, contrasting with its earlier stance on minimal AI regulation.
OpenAI has restricted the initial rollout of its latest AI models, GPT 5.6 (Sol, Terra, Luna), at the request of the US government, citing national security concerns. This marks a shift in the Trump administration’s approach to AI, from advocating minimal regulation to embracing tighter oversight.
The decision follows a similar move by Anthropic, which disabled access to its Fable 5 and Mythos 5 models after receiving a US export control directive. OpenAI has begun a limited preview for trusted partners, with plans for broader release in coming weeks.
The US Commerce Secretary noted progress with Anthropic after restrictions, though Fable 5 remains limited. Critics argue the ad hoc oversight lacks a clear legal framework, potentially reducing predictability for developers and undermining US AI competitiveness.
The administration’s stance on AI has evolved, with President Trump previously championing minimal regulation but now citing growing cybersecurity and national security concerns as capable AI systems emerge.
What to Watch
AI outlook — possibilities, not facts
Increased regulatory actions on AI companies in the US
Likely · Within weeks
Open Questions
- What specific security concerns prompted the US government's request?
- How will the new regulatory framework impact AI development globally?
