White House Tells OpenAI to Keep GPT-5.6 on a Leash While AI Watchdogs Circle 🤖

The Trump administration has asked OpenAI to limit the initial release of its GPT-5.6 model to a small group of government-approved partners while federal officials evaluate the system, according to reports by The Information and Axios. The request, made through the White House's Office of the National Cyber Director and Office of Science and Technology Policy, comes as the administration develops a framework for assessing advanced AI models before wider deployment.

The move marks the second time this month that the U.S. government has intervened to restrict a frontier AI release. Earlier in October, Washington ordered Anthropic to suspend public access to Claude Fable 5 and Mythos 5 over national security concerns, prompting the company to pull the more capable variants. Sources familiar with the discussions said the OpenAI request was driven by GPT-5.6's "Mythos-like" capabilities rather than a broader shift in AI policy, with both the administration and OpenAI's developers viewing the new model as "on par" with Mythos, per The Information.

OpenAI has agreed to the limited rollout as a way to navigate regulatory uncertainty and allow some form of public deployment to proceed. CEO Sam Altman told employees in a memo on Thursday that the government is giving the company "access" under the new approval system, adding that the arrangement does not reflect OpenAI's long-term strategy. The company has not disclosed which partners will receive access or how long the restrictions will remain in place.

The intervention follows President Trump's executive order earlier this month directing federal agencies to establish a voluntary testing framework for advanced AI systems before release, a program shaped by weeks of internal debate over its structure. It also reflects a shift in the relationship between leading AI developers and Washington after years of industry calls for clearer regulation. During Senate testimony in 2023, Altman urged lawmakers to create a dedicated regulatory agency for advanced AI, while Anthropic CEO Dario Amodei has more recently argued that the most capable models should undergo rigorous government-backed evaluations before deployment because of potential risks tied to cyberattacks and biological weapons research.

Anthropic is also pressing Congress to strengthen protections against AI model distillation, alleging in a June 10 letter to Senate Banking Committee Chairman Tim Scott and Ranking Member Elizabeth Warren that operators affiliated with Alibaba and its Qwen AI lab generated more than 28.8 million exchanges with Claude between April and June in the largest known effort to extract capabilities from the chatbot. OpenAI, Anthropic and Google have each published proposals outlining how frontier AI should be governed, with all three calling for structured evaluations of the most capable models, greater transparency around safety testing, independent review of high-risk systems and a larger government role in overseeing development, though their approaches differ.