OpenAI, the creator of ChatGPT, announced on Friday that it is limiting the rollout of its latest artificial intelligence model following a request from the U.S. President Donald Trump’s administration. This move comes amidst heightened government scrutiny of AI technologies that may pose cybersecurity threats.
The new AI model, named GPT-5.6 Sol, will initially be accessible only to a select group of approved partners designated by the Trump administration. OpenAI emphasized that it does not consider this government access process as a permanent solution and views it as a temporary measure before making the model more widely available in the near future.
This step by OpenAI to gradually introduce its powerful new AI system mirrors recent actions taken by the government against its competitor, Anthropic, the developer of the Claude chatbot. Anthropic swiftly took down its latest AI models, Fable 5 and Mythos 5, shortly after their public release to comply with a directive barring their use by non-U.S. citizens.
The White House affirmed its ongoing collaboration with leading AI research labs to address the challenges associated with advancing AI technologies. Concerns have escalated following Anthropic’s warning earlier this year about the potential weaponization of its Mythos model to exploit software vulnerabilities, posing a threat to global computer networks.
In June, President Trump issued an executive order mandating a 30-day national security review of advanced AI systems before their public launch. Although participation by AI developers is voluntary, the framework for this review process is still in development.
OpenAI clarified that its new Sol model primarily focuses on identifying and addressing vulnerabilities rather than engaging in cyberattacks, staying within the company’s risk tolerance threshold. However, the company acknowledged the possibility of unforeseen risks, especially in combination with other tools, prompting a cautious approach to its release.
The decision to implement a phased rollout of the Sol model stems from the uncertainties surrounding its enhanced capabilities and the need for robust safeguards. Cybersecurity experts have voiced criticism over the government’s actions that led Anthropic to deactivate Fable, a version promoted as a safer alternative to Mythos, which has been offline for two weeks.
Alex Stamos, a cybersecurity expert from Stanford University, questioned the justification behind shutting down Fable, noting that industry experts do not see any concrete basis for this action. Stamos highlighted that such moves could hinder the U.S.’s competitive edge against China in the AI race.
In conclusion, OpenAI’s decision to restrict the release of its GPT-5.6 Sol model aligns with heightened government oversight of AI technologies for national security reasons, reflecting a broader trend of regulatory scrutiny in the tech industry.