OpenAI says it is limiting the release of its latest ChatGPT-related AI model, GPT-5.6 Sol, while a cybersecurity review is underway at the request of President Donald Trump’s administration. According to the reports, OpenAI is not making the model broadly available at this stage and instead is sharing it only with a limited set of “trusted partners.” The outlets describe the partners as being approved by the Trump administration, framing the restriction as part of the government’s review process for security concerns. The company characterizes the restriction as temporary and tied to the ongoing cybersecurity assessment. The Independent and Japan Today both report that the limitation is being imposed by the administration, while Stuff.co.nz similarly states that access is restricted to a small group of partners that the administration approves. Across the accounts, the central points are that GPT-5.6 Sol’s wider release is paused, access is limited to approved partners, and the stated reason is the cybersecurity review. No additional technical details about the model or the review’s scope are provided in the excerpts.