arrow_backBack to Blog
OpenAIGPT-5.6regulationWhite Housesecurity

White House limits OpenAI GPT-5.6 release to government-approved partners

personLab Desenvolvimento
schedule5 min

The US government under the Trump administration has requested that OpenAI restrict the launch of GPT-5.6 to government-approved partners. The reason? National security concerns around a model that sources say is comparable to Anthropic’s Mythos in capability.

Sam Altman, OpenAI’s CEO, stated this is ‘the best path forward’ to release the model, signaling the company prefers cooperating with the government rather than facing stricter blocks. The decision reflects a tense moment for the AI industry: increasingly powerful models are raising alarms worldwide.

GPT-5.6 had been in internal testing for months, with expectations of a public release this year. Now access will be controlled, with government and security partners getting priority. Regular companies and developers may have to wait longer.