OpenAI is facing fresh scrutiny from the White House as the Trump administration has reportedly asked the company to stagger the release of its next model because of security concerns. The request, first reported by The Information, centers on GPT-5.6 and reflects growing government attention on the risks associated with rapid AI deployment.

According to the report, administration officials want OpenAI to slow the rollout rather than make the model broadly available at once. The aim appears to be giving security teams more time to assess potential risks before the model reaches a wider set of users. The publication did not detail the specific threats that prompted the concern.

The reported intervention underscores how advanced AI systems are increasingly becoming a matter of national policy, not just product strategy. As models become more capable, governments in the United States and elsewhere have raised questions about misuse, safety testing, and the possibility of systems being deployed before companies have fully evaluated their behavior.

OpenAI has not publicly described any change to its release plans in response to the report. The company has, over time, faced pressure from regulators, lawmakers, and safety advocates to move carefully as it introduces more powerful versions of its models. Staggered launches can give developers a chance to monitor early use, identify unexpected failures, and limit exposure if problems emerge.

The reported request also comes at a moment when the AI industry is racing to ship newer models at a fast pace. That competition has made release timing a strategic issue for leading companies, but it has also intensified debate over whether faster deployment is compatible with adequate oversight. If the administration’s reported concerns are accurate, they suggest the government is pushing for a more cautious approach even as commercial incentives point in the opposite direction.

The Information’s report does not say whether the White House’s request was formal or informal, nor does it specify how OpenAI responded. It also does not indicate whether the model’s launch would be delayed, limited to certain users first, or released in phases.

Still, the episode highlights a broader tension in the AI sector. Companies are under pressure to deliver increasingly powerful systems, while policymakers are seeking ways to ensure those systems are introduced without creating avoidable security or safety risks. A staggered release, if adopted, would fit that balancing act by allowing more observation before full-scale availability.

For now, the reported request places OpenAI at the center of another public debate over how quickly frontier AI should advance and how much control governments should have over that pace.