President Donald Trump, who previously advocated for minimal government intervention in the realm of artificial intelligence, is now contemplating the implementation of regulatory oversight for emerging AI technologies, as reported by U.S. officials and sources familiar with the discussions.
The administration is exploring the possibility of an executive order to establish a working group focused on AI, which would gather technology leaders and government representatives to assess potential oversight mechanisms. Officials, who requested anonymity to discuss sensitive matters, indicated that one of the main proposals under consideration is a formal review process for new AI models.
During discussions held last week, White House representatives shared some of these plans with executives from major tech firms including Anthropic, Google, and OpenAI, according to insiders present at the meetings.
The proposed working group is expected to evaluate various oversight strategies. Some officials suggested that the review process may mirror efforts currently being developed in the United Kingdom, where multiple government agencies are tasked with ensuring that AI technologies adhere to specific safety standards, as noted by participants from both the tech industry and the administration.
This development marks a significant shift in the Trump administration’s stance on AI. Since his return to office last year, Trump has been a strong advocate for the technology, asserting its importance in the geopolitical rivalry with China. He quickly dismantled regulatory measures set by the Biden administration that required AI developers to conduct safety assessments and disclose information on AI systems with potential military implications.
At a July event, Trump expressed his enthusiasm for AI, stating, “We’re going to make this industry absolutely the top, because right now it’s a beautiful baby that’s born. We have to grow that baby and let that baby thrive. We can’t stop it. We can’t stop it with politics. We can’t stop it with foolish rules and even stupid rules.” He acknowledged the need for regulations but emphasized that they must be more innovative than the technology itself.
Despite his enthusiasm, Trump appears to be increasingly isolated in his views on AI. With rising public concern regarding the potential risks of AI—impacting jobs, energy costs, education, privacy, and mental health—there is growing bipartisan agreement. A Pew Research Center survey conducted last year revealed that 50% of Republicans and 51% of Democrats expressed more apprehension than excitement about the expanding role of AI in everyday life.
This shift in policy was triggered in part by Anthropic’s announcement of a powerful new AI model named Mythos, which the company claims could revolutionize cybersecurity by identifying vulnerabilities in software. Anthropic has opted not to release this model publicly due to its capabilities.
The White House aims to mitigate political fallout in the event of a significant AI-related cyber incident, according to insiders. Additionally, the administration is examining the potential military applications of new AI models for the Pentagon and U.S. intelligence agencies. Some officials are advocating for a review system that would allow the government early access to AI models while still permitting their public release.
This evolving stance on AI has created uncertainty within the industry. As discussions between the White House and tech executives progress, some leaders have cautioned that excessive government oversight could hinder American innovation in the face of competition from China. However, there is also a lack of consensus among companies regarding the best path forward for regulation.
Dean Ball, a former senior adviser on AI in the Trump administration, remarked, “The technology is moving extremely fast, and there are few formal procedures, but they also don’t want to overregulate. It’s a tricky balance.”
A White House official clarified that any discussions about a potential executive order remain speculative, and that Trump would personally announce any policy decisions.
These changes in AI policy coincide with a leadership transition within the White House. In March, David Sacks, the AI czar who had led deregulation efforts, announced his departure. Susie Wiles, the White House chief of staff, and Treasury Secretary Scott Bessent have since assumed responsibility for AI policy, indicating their intention to have a more significant influence in this area.
However, Wiles and Bessent’s efforts have been complicated by an ongoing dispute between Anthropic and the Pentagon. The two parties are currently at odds over a $200 million contract and the application of AI in military operations. Following a failure to reach an agreement, the Pentagon suspended its use of Anthropic’s technology earlier this year, prompting Anthropic to file a lawsuit against the government.
This conflict has posed challenges for certain government agencies that have come to depend on Anthropic’s technology. Notably, Anthropic’s AI is still utilized in the military’s Maven system, which assists in intelligence analysis and targeting for airstrikes in Iran.
Recently, the National Security Agency has also employed Anthropic’s Mythos model to evaluate vulnerabilities in the software used by the U.S. government, according to knowledgeable sources.
In a recent meeting at the White House, Wiles and Bessent engaged with Anthropic CEO Dario Amodei to discuss reinstating the company’s technology for government use. Both parties characterized the meeting as “productive.”
Officials indicated that should the administration proceed with the vetting of AI models, the working group would play a crucial role in determining the relevant agencies involved in this oversight process.
