OpenAI, the developer of ChatGPT, announced on Friday that it is limiting the launch of its latest artificial intelligence model in response to a request from the U.S. President Donald Trump’s administration. This move reflects the government’s unprecedented scrutiny of AI technologies that could potentially pose cybersecurity threats.
The new AI model, named GPT-5.6 Sol, will initially be accessible only to a select group of trusted partners who have been approved by the Trump administration. OpenAI emphasized that this restricted government access should not become the standard practice in the long term.
The company described this restricted release as a temporary measure on the path towards broader availability in the upcoming weeks. This decision comes after the government’s recent actions against Anthropic, a competitor of OpenAI known for its Claude chatbot. Anthropic voluntarily 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 foreign individuals.
The White House confirmed ongoing collaboration with leading AI labs to address the challenges associated with the rapid expansion of AI technology. Concerns had escalated following Anthropic’s warning earlier this year about the potential weaponization of its Mythos model by hackers to exploit software vulnerabilities, posing risks to global computer networks.
President Trump signed an executive order in June focusing on AI oversight, establishing a framework for evaluating the national security implications of advanced AI systems for up to 30 days before their public launch. OpenAI assured that its new Sol model primarily focuses on aiding in identifying and addressing vulnerabilities rather than engaging in cyberattacks. However, the company acknowledged potential unforeseen risks, especially when the model is combined with other tools.
To mitigate uncertainties and enhance safety measures, OpenAI announced a phased release strategy for its advanced AI model. Cybersecurity experts have criticized the government’s actions that led to the shutdown of Anthropic’s Fable model, which was positioned as a safer variant of Mythos. The model has been offline for two weeks, drawing skepticism from industry experts regarding the factual basis for the government’s intervention.
Alex Stamos, a cybersecurity expert from Stanford University, emphasized that the cybersecurity community largely perceives the government’s actions as unfounded. Stamos highlighted that the decision could hinder the U.S.’s competitive edge against China in the AI race. He noted that the risks associated with Fable are comparable to those of publicly available AI models, including those developed in China.