OpenAI Reportedly Restricts GPT-5.6 Preview to Government-Approved Partners

OpenAI is reportedly preparing to launch its next-generation AI model, GPT-5.6, but unlike previous releases, access may initially be limited to a small number of government-approved partners.

According to The Information, the company will not conduct a broad public rollout during the preview phase. Instead, access to GPT-5.6 will reportedly be granted on a customer-by-customer basis with approval from the U.S. federal government.

OpenAI CEO Sam Altman reportedly told employees that this restrictive deployment is intended only for the initial preview period. He also reportedly said the company does not see this as its long-term approach to distributing AI models and hopes to move toward a more sustainable deployment model in the future.

Government review could shape the rollout

The reported launch strategy comes as the U.S. government increases its focus on advanced AI systems.

According to CNN, both OpenAI and the Trump administration reportedly view GPT-5.6 as having a similar capability level to Anthropic’s Mythos AI models.

President Donald Trump recently signed an executive order asking AI companies to voluntarily submit their most powerful AI models for federal review before making them publicly available.

If implemented as reported, GPT-5.6 would become one of the first major AI models to undergo such a controlled preview process.

Security concerns continue to grow

The reported restrictions follow the release of Anthropic’s Mythos 5, which was later shut down abruptly. According to the information provided, the model was reportedly able to identify vulnerabilities in NSA systems during testing.

At the same time, Anthropic has reportedly accused Alibaba of attempting to distill its AI technology to create its own version of Mythos 5.

These developments have added to growing concerns surrounding frontier AI models, particularly regarding cybersecurity, intellectual property, and national security.

While OpenAI has reportedly described the government-controlled rollout as a temporary measure, the preview of GPT-5.6 may signal a broader shift toward closer government oversight of the industry’s most capable AI systems.

Via Engadget