OpenAI restricts new model access after Trump administration security review request
In brief
- Trump administration requests OpenAI and Anthropic restrict advanced model rollouts
- OpenAI launches GPT-5.6 Sol, Terra, Luna on June 26 with limited government-approved access
- Executive order mandates 30-day pre-release security reviews for advanced AI models
- OpenAI warns federal vetting could stifle innovation and slow development cycles
Government vetting enters the AI race
The Trump administration requested that OpenAI and Anthropic limit model rollouts, citing national security concerns. Before making its announcement, OpenAI previewed the models' capabilities to government officials. Those officials, focused on cybersecurity and national security implications, then requested the company restrict who gets early access.
An executive order from the Trump administration encourages a 30-day pre-release security review by AI companies before launching advanced models. OpenAI cooperated this time, accepting restricted access and a review period before broader rollout.
The innovation concern
OpenAI has stated that routine federal vetting of AI releases would stifle innovation, slow development cycles, and hurt users, developers, and enterprises that depend on rapid access to new tools. The company's position reflects a broader tension between security oversight and competitive speed in the AI sector.
Advanced AI models have demonstrated capabilities in code generation, vulnerability discovery, and persuasion that concern cybersecurity professionals. These capabilities explain why officials requested restrictions, though the timeline and scope of such reviews remain unclear.
Competitive implications
Anthropic faced similar directives regarding its Mythos model. If the federal government can effectively delay model launches through security reviews, then the competitive dynamics of the AI race change meaningfully. OpenAI has indicated broader access to Sol, Terra, and Luna will follow once the review period concludes, but the precedent of pre-release vetting may reshape how the industry operates going forward.
Frequently asked questions
Why did the Trump administration ask OpenAI to restrict model access?
Government officials focused on cybersecurity and national security implications requested that OpenAI limit early access to its new models. Advanced AI models have demonstrated capabilities in code generation, vulnerability discovery, and persuasion that concern cybersecurity professionals.
How long will the security review take?
An executive order from the Trump administration encourages a 30-day pre-release security review by AI companies before launching advanced models. OpenAI has indicated broader access will follow once the review period concludes.
Does this affect other AI companies?
Anthropic, one of OpenAI's primary competitors, faced similar US government directives regarding its Mythos model, suggesting the policy applies across the industry.
