OpenAI narrows GPT-5.6 access

OpenAI said Friday it is limiting the rollout of its newest GPT-5.6 models to a “small group of trusted partners” after a request from the U.S. government, underscoring growing tension between AI developers and federal officials over how advanced systems should be released.

The GPT-5.6 lineup includes Sol, OpenAI’s flagship model; Terra, a more balanced option for everyday use; and Luna, a faster, lower-cost version. OpenAI said the preview has been restricted to partners “whose participation has been shared with the government.”

The company said the move is temporary, but it also made clear it does not want this kind of oversight to become standard practice. “We don’t believe this kind of government access process should become the long-term default,” OpenAI wrote in a Friday blog post. “It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.”

Pressure from Washington

The administration’s request comes as the U.S. government increases pressure on AI companies to limit access to their most advanced systems. After Anthropic launched its most powerful public model, Fable 5, the administration ordered the company to remove access for any foreign national, prompting Anthropic to take the model down entirely.

That episode has fueled debate over how much authority the government should have over model releases. Dean Ball, a former White House AI adviser and soon-to-be OpenAI employee, said President Trump’s recent executive order — which asks some AI companies to voluntarily submit their most advanced models for government review up to 30 days before release — has effectively created a licensing regime for frontier AI.

Ball argued that without clearly defined safety standards, companies could face repeated delays that slow launches, weaken U.S. competitiveness and complicate the massive infrastructure investments pouring into AI.

OpenAI said its GPT-5.6 preview is a “short-term step” and that the model is still on track for broader availability in the coming weeks. The company said it is working with the administration on a new executive order framework focused on cybersecurity, along with a “repeatable process for future model releases.”

OpenAI says GPT-5.6 Sol is its strongest model yet, with improved agentic capabilities in coding, biology and cybersecurity.

Sources:

techcrunch.com