The administration of US President Donald Trump has asked OpenAI to postpone the full launch of the new artificial intelligence model GPT-5.6. Initially, access to the model is planned to be granted only to a limited circle of verified partners, reports UNN citing Bloomberg.
Details
According to the publication, OpenAI CEO Sam Altman told employees that the US government proposed first testing GPT-5.6 among about 20 trusted partners, and only then opening wider access. The model is expected to be available through the Amazon Bedrock platform.
OpenAI does not officially comment on this information. At the same time, the White House stated that the administration continues to cooperate with leading artificial intelligence developers to develop joint approaches to the safe implementation of new technologies.
Washington is concerned about the capabilities of new AI models
During a meeting with employees, Altman said that US authorities are increasingly concerned about the capabilities of the most advanced artificial intelligence models. He also urged the team to cooperate with the government on possible safety requirements and restrictions, even if the company does not share certain decisions of officials.
According to Bloomberg, a similar situation has already arisen with the company Anthropic, which at the beginning of the month was forced to restrict access to its most powerful models after US authorities intervened due to national security considerations.
The US is preparing new rules for AI developers
According to the agency, the Trump administration is working to create a unified system of interaction between the government and companies developing advanced artificial intelligence models. The corresponding executive order stipulates that within 60 days, the government and industry representatives must prepare a voluntary mechanism that will allow US authorities to gain access to new models approximately 30 days before their official launch.
The industry expects that the new rules will make the approval process more transparent, since currently companies do not have clear instructions on when and to what extent to notify the government about the development of new artificial intelligence models.