OpenAI GPT-5.6 Sol release restricted after US cybersecurity vetting request

OpenAI said it is limiting access to its new artificial intelligence model after a request from President Donald Trump’s administration. The company said the model, called GPT-5.6 Sol, is being released only to a small set of trusted partners. OpenAI said those partners were approved through the administration’s review process.

Representative image

OpenAI said the restriction is not meant to be permanent. "We dont believe this kind of government access process should become the long-term default,\" OpenAI said in a statement. The company said the current testing phase is temporary. OpenAI added that wider access is expected over the coming weeks.

Trump administration AI oversight and GPT-5.6 Sol release

The White House said it is working with frontier AI labs on problems linked to scaling the technology. Earlier in June, Trump signed an executive order on AI oversight. It set a process to assess national security risks. Under the order, advanced AI systems could face up to 30 days of review.

The executive order described developer participation as voluntary. However, the full framework is still not complete. Even so, the new process has already shaped how companies release models. OpenAI’s phased rollout followed the administration’s emphasis on cybersecurity and national security checks.

OpenAI GPT-5.6 Sol safeguards and cybersecurity risk concerns

OpenAI said Sol performs better at helping users spot and fix software weaknesses than launching cyberattacks. The company said the model does not cross OpenAI’s own risk threshold. OpenAI also said risks could appear in unexpected ways. It warned this could happen if Sol is combined with other tools.

\"That uncertainty, along with the models broader step change in capabilities, is why we are pairing the models increased capabilities with stronger safeguards and a phased release,\" the company said Friday. OpenAI said the approach balances stronger protections with plans for broader availability. OpenAI did not list the partners with access.

Anthropic Claude models and Trump directive on foreign nationals

OpenAI’s rollout followed actions taken earlier this month involving Anthropic, which makes the Claude chatbot. Anthropic took offline two models, Fable 5 and Mythos 5, days after releasing them publicly. The company said it acted to meet a Trump directive. The directive blocked use by foreign nationals.

Government concern increased after Anthropic warned earlier this year about Mythos. Anthropic said Mythos could find software flaws in ways that might be weaponised. Officials said such abilities could threaten critical computer networks worldwide. That warning added to White House scrutiny of powerful, newer AI systems.

Cybersecurity experts criticise Trump administration AI actions

Many cybersecurity experts criticised the government steps that led to Fable being shut down. Anthropic had described Fable as safer than Mythos. The model has now been unavailable for two weeks. Some critics said the move could weaken US competitiveness and disrupt research and security work.

\"I just want to say that pretty much nobody in the cybersecurity industry believes that theres any factual basis for this action,\" Stanford University cybersecurity expert Alex Stamos said on a call with reporters earlier this week. Stamos is also chief product officer at AI security company Corridor and a former Meta security officer.

Stamos said he reviewed research analysis by Amazon, Anthropic’s main cloud backer. Stamos said the review did not show risks beyond those seen in other public models. Stamos said this included models made in China. \"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,\" Stamos said.

Trump technology advisers and OpenAI talks with Howard Lutnick

Some Trump allies blamed Anthropic and CEO Dario Amodei for increased scrutiny. \"Dario came to Washington a few months ago, back in April, and basically said that he had created a cyber weapon called Mythos, said investor David Sacks, who co-leads Trumps council of technology and science advisers, on a recent podcast. And he spiked the cortisol level, got everyone really worried. And there was some truth to it in terms of the sense that this model had advanced cyber capabilities.\"

OpenAI CEO Sam Altman spoke with US Commerce Secretary Howard Lutnick about the release on Wednesday. The call was part of wider talks between AI leaders and Trump officials in recent weeks. Anthropic also joined discussions, though Amodei’s relationship with the administration has been more tense.

The Pentagon designated Anthropic as a national security risk after the company raised ethical and safety concerns about AI in war. Trump also ordered federal agencies to stop using Claude. Anthropic responded with a lawsuit that is still moving through federal courts. For OpenAI, the company said Sol access remains limited while review continues.

With inputs from PTI