tech
OpenAI to Restrict GPT-5.6 Release at U.S. Government's Request
At the request of the Trump administration, OpenAI will limit the initial release of its new GPT-5.6 models, including Sol, Terra, and Luna, to a select group of government-approved partners. The staggered rollout is intended to allow for user vetting and security reviews due to the models' powerful capabilities, marking a significant instance of government oversight in AI deployment.
TL;DR
- The U.S. government has requested OpenAI to stagger the release of its new GPT-5.6 models, Sol, Terra, and Luna.
- This restricted preview is limited to approximately 20 "trusted partners" pre-cleared by the U.S. government.
- The decision stems from an executive order asking leading AI firms to voluntarily submit advanced models for government review before release.
- The government cited "Mythos-like" capabilities in cybersecurity and other sensitive domains as a reason for the staggered rollout.
- OpenAI CEO Sam Altman stated this is not their preferred long-term model for launches but complied with the government's request.
- The move is seen as a test case for the government's AI oversight and has raised concerns about potential "de facto involuntary licensing."