OpenAI's Restricted GPT-5.6 Rollout Amid US Security Concerns

OpenAI's rollout of its GPT-5.6 model has become a focal point of tension between the tech giant and the US government. On June 25, 2026, OpenAI CEO Sam Altman informed staff that the White House requested the company to restrict access to the new model during its initial preview phase. This request was driven by security concerns, leading to a decision to grant access on a customer-by-customer basis, mirroring similar actions taken by competitors like Anthropic with its Mythos model.

The following day, reports confirmed that OpenAI would limit access to GPT-5.6 to a select group of about 20 government-approved partners. This staggered approach aimed to address the administration's concerns while still allowing for some level of testing and feedback. By June 27, the limited preview officially began, with access granted only through API and Codex, underscoring the tightly controlled nature of this rollout.

The stakes are high as this situation unfolds. OpenAI's ability to innovate and release new models is now intertwined with governmental oversight. This could set a precedent for how AI technologies are governed in the future, potentially stifling rapid advancements in the field. As the industry watches closely, the implications of this restricted rollout could lead to broader conversations about AI safety, ethics, and the balance between innovation and regulation.