Government access first
OpenAI announced it would hold back the public release of GPT-5.6 after the Trump administration requested priority access for federal agencies. The U.S. government is seeking to deploy the model across national security, intelligence, and defense applications before it becomes broadly available.
OpenAI pushes back on precedent
While complying with the request, OpenAI's leadership made clear they oppose making government-first access a standard practice. 'These restrictions should not become the norm,' the company stated, warning that repeated delays to public releases could undermine transparency and slow innovation in the broader AI ecosystem.
Frontier AI control debate intensifies
The standoff mirrors similar tensions with Anthropic, whose Mythos model was recently cleared for release to over 100 U.S. companies and agencies after months of export restrictions. The pattern suggests the U.S. government is asserting greater control over cutting-edge AI models, treating them as strategic assets akin to semiconductor technology.