Trump and AI: Towards a Regulation that Divides Experts

Trump and AI: Towards a Regulation that Divides Experts

The administration of Donald Trump is considering implementing oversight of new artificial intelligence (AI) models, according to reports from the New York Times. This initiative has sparked mixed reactions among technology policy experts, who fear it may stifle innovation. They emphasize that any regulation should be the result of legislation rather than a presidential decree.

The New York Times report indicates that the Trump administration is currently discussing the formation of a task force. This group would include leaders from the tech industry as well as public officials, with the aim of defining the terms of this oversight. White House representatives have already begun informing companies like Anthropic, Google, and OpenAI about these plans.

This move would represent a significant shift from Trump’s initial policy on AI. During his second term, the president had emphasized deregulation and encouragement of innovation, arguing that overly strict controls could weaken the competitiveness of the United States against China.

Pressures and Concerns Surrounding AI

This initiative comes at a time when policymakers are under pressure. National security officials and technology analysts are urging them to consider the risks associated with powerful AI systems, such as the Mythos model developed by Anthropic.

Divergent Opinions Among Experts

• Daniel Castro, president of the Information Technology and Innovation Foundation (ITIF), expressed his opposition to a potential executive order. He considers this idea a "terrible idea," stressing that it would embody a "total acceptance" of the precautionary principle. This would require companies to obtain government approval to innovate, slowing the pace of innovation from that of Silicon Valley to that of Washington.

• Janet Vestal Kelly, CEO of the Alliance for a Better Future, welcomed the idea of verifying AI models. She stated that AI is the most powerful technology ever developed and that without regulation, large tech companies could harm children, workers, and American values.

• Adam Thierer, an analyst at the R Street Institute, warned that AI regulation should not be a unilateral decision by Trump. He cautioned against a "preemptive verification" by the White House, which could turn into a de facto licensing regime.

• Conor Grennan, an AI consultant and former chief AI architect at NYU's Stern School of Business, expressed concerns about the verification process. He raised the issue of accountability in case of problems and warned that government decisions could quickly become complex.

• Eli Dourado, head of strategic investments at the Astera Institute, noted that this proposal could conflict with the prior restraint principle of the First Amendment, which prohibits the government from preventing speech before it occurs.

• Taylor Barkley, director of federal government affairs at the Abundance Institute, compared the potential regulation to a pre-approval process similar to that in the UK, which would represent a "huge step backward for innovation."

• Chris McGuire, senior fellow at the Council on Foreign Relations, described this move as a "necessary regulatory pivot." He emphasized the importance of establishing safety requirements for cutting-edge AI labs and cloud service providers.

• Thomas Woodside, senior advisor at the Secure AI Project, expressed doubts about the effectiveness of a single approval process for AI models. He advocated for ongoing oversight, as systems are frequently updated and can pose risks even during their development.

Experts agree on one point: any regulatory framework should be established through legislation rather than executive action, to ensure a balanced and sustainable approach.