In a dramatic shift in US technology policy, President Donald Trump has signed an executive order repealing a 2023 directive that required AI companies to report safety data to the federal government. This decision, executed on January 20, 2025—Trump’s first full day in office—signals a significant departure from the regulatory approach established under President Joe Biden.
The Rescinded AI Safety Order
The original executive order, signed by Biden in 2023, marked the most comprehensive federal regulation of AI development in US history. It mandated that companies developing AI systems with potential implications for national security, public safety, health, or the economy submit safety and risk assessment data before public deployment. Enforced under the Defense Production Act, it also tasked federal agencies with establishing safety standards to mitigate risks, including cybersecurity threats and chemical, biological, radiological, and nuclear hazards.
Trump’s repeal aligns with the 2024 Republican Party platform, which criticized the regulations as barriers to AI innovation. This rollback eliminates mandatory transparency and safety requirements for AI developers, raising questions about the future role of the US AI Safety Institute, a Commerce Department initiative focused on voluntary AI standards.
Global AI Policy Context
The US’s move contrasts sharply with Europe’s AI regulatory strategy. In 2024, the European Union implemented the AI Act, a robust framework that restricts certain applications, such as facial recognition in public spaces, and imposes strict oversight on high-risk AI uses in sectors like healthcare and law enforcement. While the US shifts toward a deregulated approach, nations worldwide are crafting their own AI governance frameworks to address the rapid advancements in artificial intelligence.
State-Level AI Laws and Industry Reactions
Despite the federal rollback, state-level AI laws remain in effect. California, a hub for AI innovation, has implemented regulations addressing AI transparency and deepfake technologies. Meanwhile, states like Colorado and Illinois have enacted measures to combat algorithmic discrimination in hiring.
The tech industry’s response to Trump’s decision has been mixed. While some welcome the reduced regulatory burden, others express concerns about the risks of unregulated AI development. Nvidia’s Vice President of Government Affairs, Ned Finkle, criticized recent restrictions on AI chip exports and warned that excessive deregulation could harm innovation and economic growth.
Future Implications for US AI Policy
Trump’s AI strategy appears to prioritize minimal government intervention, although the administration has yet to outline a replacement for the rescinded order. David Sacks, a venture capitalist and vocal critic of tech regulation, has been appointed to lead efforts on AI and cryptocurrency policy. The long-term consequences of these policy changes remain uncertain, but the global competition for AI dominance underscores the high stakes of this pivotal moment.
