The United States government has asked OpenAI to restrict the launch of its upcoming model, GPT-5.6, and limit it to a small group of partners approved by the administration. This measure responds to the advanced capabilities of the system, which raise security concerns.
According to reports, the decision did not come out of nowhere. Recently, the Trump administration imposed export controls on Anthropic, leading that company to withdraw its most powerful models, Mythos and Fable, from the market. These models had managed to breach the security of various government entities.
GPT-5.6 is at a similar level to Mythos, making it the second frontier model under direct scrutiny in just a few weeks. OpenAI accepted this restriction as a temporary solution to move forward with the public launch.
A "strange moment" for the industry
Sam Altman, the company's leader, described the situation in an internal report as a "strange moment", given the lack of a clear regulatory framework for artificial intelligence. "We have made it clear to the US Government that this is not our preferred model in the long term," he stated, adding that they will work alongside other stakeholders to seek more sustainable approaches.
