OpenAI announced Friday that it is restricting the release of its new artificial intelligence model, GPT-5.6 Sol, pending approval from President Donald Trump's administration as the federal government expands its vetting of AI products for cybersecurity risks. The company said the model would be accessible only to customers approved by the Trump administration, characterizing the testing period as a temporary step on the "path to broader availability in the coming weeks."
The move marks a significant shift in how advanced AI systems enter the market, with government pre-release review now a condition of deployment rather than a post-launch consideration. OpenAI has not identified any of the roughly 20 customers approved to use the new model so far.
The Government Framework
Trump earlier in June signed an executive order on AI oversight that established a framework for the federal government to vet the national security risks of the most advanced AI systems for up to 30 days before their public release. The order described participation by AI developers as voluntary, though the framework has not yet been fully developed.
OpenAI stated that "we don't believe this kind of government access process should become the long-term default," signaling the company's preference for a time-limited arrangement rather than a permanent regulatory structure. The company characterized its new Sol model as "better at helping people find and fix vulnerabilities" than at carrying out cyberattacks and said it does not cross the company's own risk threshold.
However, OpenAI acknowledged that "there could be unforeseen risks, especially if the model is combined with other tools," justifying the phased release approach alongside stronger safeguards.
Competitive Dynamics and Political Pressure
Anthropic, OpenAI's chief rival, announced hours after OpenAI's announcement that the Trump administration had approved a limited release of its strongest cybersecurity model, Mythos 5, two weeks after the U.S. Commerce Department effectively banned it. Anthropic said the government on Friday lifted restrictions on the model, enabling it to be "redeployed to a small group of cyber defenders and infrastructure providers."
Anthropic took offline two new AI models, known as Fable 5 and Mythos 5, earlier this month to comply with a Trump directive blocking their use by foreign nationals. The company said Friday it was "pleased" by the partial release of Mythos and will "continue to work with the government to expand access" and make Fable available again to general users. U.S. Commerce Secretary Howard Lutnick told Anthropic in a letter dated Friday that its work to address the government's concerns "yielded significant progress."
The divergent treatment of the two companies reflects underlying tensions. OpenAI CEO Sam Altman spoke with Commerce Secretary Howard Lutnick about the model release Wednesday, part of ongoing negotiations between AI industry executives and Trump officials. Anthropic has also been part of those talks, but CEO Dario Amodei has had a more contentious relationship with the Trump administration. The Pentagon designated Anthropic as a national security risk for raising ethical and safety concerns about AI usage in war, and Trump himself ordered federal agencies to stop using Claude. Anthropic responded with a lawsuit that is still working its way through federal courts.
Investor David Sacks, who co-leads Trump's council of technology and science advisers, said on a recent podcast that Amodei "came to Washington a few months ago, back in April, and basically said that he had created a cyber weapon called Mythos," and that he "spiked the cortisol level, got everyone really worried." Sacks acknowledged that "there was some truth to it in terms of the sense that this model had advanced cyber capabilities."
Industry Skepticism
Criticism of the government's approach has emerged from unexpected quarters. Stanford University cybersecurity expert Alex Stamos, chief product officer at AI security company Corridor and a former chief security officer at Facebook parent Meta, said on a call with reporters earlier this week that "pretty much nobody in the cybersecurity industry believes that there's any factual basis for this action." Stamos said he reviewed an analysis of research on Fable by Anthropic's primary cloud computing backer, Amazon, and "didn't find any risks that aren't present with other publicly available AI models, including those made in China."
Stamos stated, "If the administration is honest about wanting the United States to beat China in this race, then this is about the dumbest thing they could possibly do."
U.S. Rep. Lori Trahan, a Massachusetts Democrat and co-author of a bipartisan bill that would regulate AI, said in a statement that she is concerned "the Trump administration is deciding company by company who gets access to the newest AI model. No law. No process. No oversight. Just appointees in Washington deciding who's in and who's out."
Market and Capital Implications
The government's heightened AI oversight adds another complication to exploratory moves by OpenAI and Anthropic to take their companies public on Wall Street, following SpaceX's record-setting June 12 initial public offering. Trump has floated the possibility of the U.S. government owning a stake in leading AI companies, describing a concept where "pieces could be given to the American public, where the American public essentially becomes a partner with the companies."
The White House said Friday it continues to collaborate with frontier AI labs on addressing the challenges of scaling the fast-growing technology.
Why This Matters:
The government's assumption of pre-release vetting authority over private AI systems establishes a precedent for regulatory gatekeeping in an industry critical to American competitiveness. The framework's voluntary characterization masks the practical coercion evident in how Anthropic was effectively forced offline before receiving selective approval—a pattern that grants government officials discretionary power over market entry and competitive positioning. The divergent treatment of OpenAI and Anthropic raises questions about consistency and rule-of-law application in regulatory decision-making. Additionally, the vetting process introduces uncertainty into product timelines and capital planning for companies considering public offerings, potentially dampening investment returns and innovation incentives. The stated concern about cybersecurity risks, while legitimate, must be weighed against industry experts' assessment that comparable risks exist in publicly available models, suggesting the regulatory response may exceed the threat level. Finally, the administration's exploration of government equity stakes in private AI companies represents an unprecedented expansion of state involvement in commercial technology ventures.