OpenAI has been asked by the U.S. federal government to initially limit access to its latest AI model, GPT-5.6, to a small group of partners. CEO Sam Altman announced this during an internal Q&A session on Wednesday, according to a report by The Information. The request is believed to stem from the Trump administration's recent executive order, which called for a voluntary review of new AI models, particularly regarding cybersecurity. The government has asked OpenAI to restrict access during the preview phase, with approval determined on a customer-by-customer basis. Altman expressed hope for a broader release a few weeks later, assuming no obstacles arise. Source: thedecoder
The push for a phased release came from discussions with two government agencies: the Office of the National Cyber Director and the Office of Science and Technology Policy. In a memo, Altman stated that the company had made it clear to the U.S. government that this was not their preferred long-term model and that they would work with government and industry stakeholders to find a more sustainable approach for future releases. Despite sharing its limited release plans with senior government officials, Altman received a call from Commerce Secretary Howard Lutnick, who warned against moving forward without additional agency approvals. Source: thedecoder
The current situation is linked to Anthropic's 'Mythos' presentation in early April, where the company emphasized cybersecurity risks and planned a phased release. The U.S. government saw both opportunities and threats in the cybersecurity landscape, leading to its intervention. After Anthropic released Fable, the first public model in the Mythos class, along with specific security guidelines, the government forced the company to pull it offline. Despite prior collaboration to find security flaws in Fable, authorities raised no objections when the company announced the release. Source: thedecoder
