White House Restricts GPT-5.6 Launch Over Security Risks

The White House has stepped in to slow down OpenAI’s rollout of GPT-5.6. The new AI model won’t see a full release until it passes government scrutiny.

OpenAI will preview GPT-5.6 only to a small, select group of government-approved partners. Access will be granted “customer by customer,” according to company insiders. CEO Sam Altman told employees this limited release is the safest path forward.

Altman expects a broader launch a couple of weeks after the preview, but only if the government signs off. The White House wants to test the model’s safeguards thoroughly before letting it loose on the public.

Security concerns are driving this cautious approach. GPT-5.6 is said to match the capability of Anthropic’s Mythos model. Mythos is no ordinary AI — it autonomously identifies software vulnerabilities and can execute multi-step cyberattacks, including malware and ransomware deployment.

The government’s worry is clear. Frontier AI models like Mythos and now GPT-5.6 can outpace human cybersecurity experts in finding and exploiting software weaknesses. This power raises red flags about unintended consequences and misuse.

In response, the Trump administration previously forced Anthropic to offline its models Fable 5 and Mythos 5. This export control move aimed to block foreign nationals from accessing these potent tools. Mythos had been distributed to roughly 40 organizations through Project Glasswing before the crackdown.

President Trump also signed an executive order urging AI companies to submit models for government testing before public release. The White House is doubling down on that policy with GPT-5.6.

Altman told OpenAI staff, “We’ve made clear to the U.S. government that this is not our preferred long-term model, and will work with them and others in industry to achieve a more sustainable approach for future releases.”

The White House is coordinating with federal bodies like the Office of the National Cyber Director and the Office of Science and Technology Policy. Commerce Secretary Howard Lutnick reportedly advised against any launch without government approval.

This isn’t just cautious PR. It’s a rare moment when a government presses pause on AI rollout to wrestle with risk. Whether this strategy slows innovation or sets a new standard remains to be seen.