# US Executive Order on AI (2025)

The US Executive Order on AI (2025) was a presidential directive issued by Donald Trump that revoked previous AI safety orders and established new federal AI policies, focusing on deregulation and American competitiveness. It was later replaced by subsequent executive actions.

The US Executive Order on AI (2025) was a presidential directive signed by Donald Trump on January 23, 2025, that fundamentally reshaped federal artificial intelligence policy. The order revoked the previous administration's executive order on AI safety and security, which had been issued in October 2023, and replaced it with a new framework emphasizing deregulation, innovation, and American global leadership in AI. The directive marked a significant shift in US government approach, moving from proactive risk mitigation toward industry-led development and reduced federal oversight.

The order was formally titled "Removing Barriers to American Leadership in Artificial Intelligence" and was among the first major policy actions of Trump's second term. It directed federal agencies to review and potentially rescind existing AI regulations that were deemed overly burdensome, and it established a new AI action plan to be developed within 180 days. The order also prioritized the development of AI infrastructure, including energy resources and data centers, and called for the recruitment of AI talent into the federal workforce.

## Background and Context

The 2025 executive order emerged against a backdrop of rapid advancements in [artificial intelligence](https://www.wikiprompt.org/wiki/artificial-intelligence) technologies, particularly the proliferation of [large language models](https://www.wikiprompt.org/wiki/large-language-model) and [generative AI](https://www.wikiprompt.org/wiki/generative-ai) systems. The previous administration had implemented a comprehensive AI regulatory framework in 2023, which required safety testing of advanced AI models, mandated reporting of safety incidents, and established standards for AI risk management. These rules were developed in response to concerns about AI bias, privacy, and potential catastrophic risks.

Trump's order argued that such regulations hindered American competitiveness, particularly against nations like China. The directive cited the need to maintain US dominance in AI development and to prevent regulatory uncertainty from driving innovation overseas. It also reflected a broader ideological divide over AI governance, with the new administration favoring voluntary industry standards over mandatory federal requirements.

## Key Provisions

The executive order contained several major components. First, it revoked the October 2023 executive order and ordered all federal agencies to identify and propose the elimination of AI-related regulations that conflicted with the new policy direction. This included reviewing rules on algorithmic accountability, data privacy, and model transparency.

Second, the order established a Presidential Council on AI Innovation, composed of federal agency heads and industry representatives, to coordinate AI policy across the government. The council was tasked with developing a national AI strategy within 180 days, focusing on infrastructure, workforce, and research priorities.

Third, the order directed the Department of Energy and the Environmental Protection Agency to expedite permitting for AI data centers and associated energy infrastructure, including natural gas and nuclear power plants. This provision aimed to address the growing energy demands of AI computing, particularly for training [neural networks](https://www.wikiprompt.org/wiki/neural-network) and running inference at scale.

Fourth, the order mandated the creation of a federal AI talent recruitment program, allowing agencies to hire AI experts through streamlined processes and offering competitive salaries. It also directed the Office of Personnel Management to develop AI training programs for existing federal employees.

## Revocation and Replacement

The 2025 executive order was itself subject to subsequent changes. In July 2025, the administration issued a follow-up executive order that refined certain provisions, particularly around AI safety testing and export controls. This replacement order maintained the deregulatory thrust but added requirements for AI companies to report certain national security-related developments, particularly involving frontier models and potential bioweapon capabilities.

The replacement order also established a new AI Safety Institute within the Department of Commerce, though with a narrower mandate than the one proposed in 2023. This institute was tasked with developing voluntary testing standards and facilitating information sharing between government and industry, rather than enforcing mandatory compliance.

## Industry Response

Major AI companies and technology firms generally welcomed the 2025 executive order. [OpenAI](https://www.wikiprompt.org/wiki/openai), [Anthropic](https://www.wikiprompt.org/wiki/anthropic), and [Google DeepMind](https://www.wikiprompt.org/wiki/google-deepmind) issued statements praising the focus on innovation and the removal of regulatory barriers. Industry groups argued that the previous administration's rules were duplicative and created compliance costs that disproportionately affected smaller companies.

However, some technology leaders expressed mixed views. Executives from [Microsoft](https://www.wikiprompt.org/wiki/microsoft) and [Amazon Web Services](https://www.wikiprompt.org/wiki/amazon-web-services) privately raised concerns about the lack of federal safety oversight, particularly given the rapid deployment of [generative AI](https://www.wikiprompt.org/wiki/generative-ai) systems in critical sectors like healthcare and finance. Several AI researchers, including [Yoshua Bengio](https://www.wikiprompt.org/wiki/yoshua-bengio) and [Geoffrey Hinton](https://www.wikiprompt.org/wiki/geoffrey-hinton), publicly criticized the order, warning that deregulation could lead to unsafe AI deployment.

## Impact on Federal AI Research

The order affected federal AI research programs, including those at [MIT CSAIL](https://www.wikiprompt.org/wiki/mit-csail), [Stanford AI Lab](https://www.wikiprompt.org/wiki/stanford-ai-lab), and [Berkeley AI Research](https://www.wikiprompt.org/wiki/berkeley-ai-research). While the order did not cut research funding, it redirected priorities toward applied AI and national security applications. The Defense Advanced Research Projects Agency (DARPA) received increased funding for AI-related projects, while civilian research programs saw more modest growth.

Federal agencies were also directed to favor procurement of commercially available AI systems over developing custom solutions. This shift benefited companies like [NVIDIA](https://www.wikiprompt.org/wiki/nvidia), [AMD](https://www.wikiprompt.org/wiki/amd), and [Intel](https://www.wikiprompt.org/wiki/intel), which supply the hardware for AI computing. The order also encouraged federal use of cloud services from providers like [Azure](https://www.wikiprompt.org/wiki/azure), [Google Cloud](https://www.wikiprompt.org/wiki/google-cloud), and [Oracle Cloud](https://www.wikiprompt.org/wiki/oracle-cloud).

## International Reactions

The executive order drew international attention, with the European Union expressing concern about the divergence in AI governance approaches. EU officials noted that the US move toward deregulation could complicate transatlantic cooperation on AI standards, particularly regarding data protection and algorithmic transparency. The United Kingdom and Japan indicated they would monitor the US approach but maintained their own regulatory trajectories.

China's response was muted, with state media framing the order as an acknowledgment of China's AI progress. The order's emphasis on American competitiveness was seen as a response to China's substantial investments in AI research and development, including work at institutions like [Alibaba Damo Academy](https://www.wikiprompt.org/wiki/alibaba-damo-academy) and [Baidu](https://www.wikiprompt.org/wiki/baidu).

## Criticisms and Controversies

Critics of the 2025 executive order raised several concerns. Consumer advocacy groups argued that revoking safety testing requirements could lead to biased or harmful AI systems being deployed without adequate oversight. Civil liberties organizations worried about the expansion of AI use in law enforcement and national security without corresponding privacy protections.

Labor unions expressed concern about the order's impact on workers, particularly in sectors likely to be automated. The order's emphasis on rapid AI adoption was seen as potentially accelerating job displacement without providing adequate retraining programs. Some economists noted that the order's deregulatory approach could exacerbate existing inequalities in the AI workforce.

## Legacy and Significance

The 2025 executive order represented a significant pivot in US AI policy, marking the first time a presidential directive explicitly prioritized deregulation over safety in the AI domain. Its effects were felt across the AI ecosystem, from research funding to corporate compliance practices. The order's emphasis on infrastructure and energy also highlighted the growing importance of computing resources in AI development.

The order's revocation and replacement demonstrated the volatility of AI policy in the face of changing political administrations. It also underscored the ongoing debate over how to balance innovation with safety in a rapidly evolving technological landscape. As of late 2025, the order's long-term impact on AI development and governance remained uncertain, with ongoing litigation and legislative efforts seeking to address some of its more controversial provisions.

---
Source: https://www.wikiprompt.org/wiki/us-executive-order-ai-2025
License: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/)
Last updated: 2026-09-12T16:23:05.47592+00:00