CSET's ETO AGORA platform provides detailed summaries, full text and thematic tags for over 900 AI-focused laws and policies from the U.S. and around the world. We added 18 documents to AGORA over the last four months. One of these documents is the AI Action Plan, the Trump administration's strategy to spur AI innovation; expand data centers, energy infrastructure, and semiconductor manufacturing facilities; and lead in international AI diplomacy and security. The Plan, which favors a de-regulatory approach to AI, was released in July 2025 following the Trump administration's revocation of Biden's expansive AI executive order.

Given that four months have passed since the Plan was released, we asked several CSET experts to comment on the Plan's policy ramifications and any actions taken so far to implement its recommendations.

⭐️ Expert Commentary

Owen Daniels, Associate Director of Analysis

The AI Action Plan is a strategic document, far-ranging in scope, that lays out the Trump administration's ambitious agenda of more than 90 recommendations across three pillars: accelerating AI innovation, building American AI infrastructure, and leading in AI security and diplomacy. Broadly, these recommendations focus on prioritizing American-made AI, both at home and abroad. Domestically, goals include building data centers, cultivating the AI and data center workforce, and building research infrastructure, while simultaneously exporting full American AI stacks abroad. The plan aims to remove red tape in the AI industry while creating incentives and easing permitting processes for building critical AI infrastructure.

Nearly four months on from the release of the AI Action Plan, implementation details remain to be seen. Three executive orders released around the same time as the plan suggest data center creation, reducing bias, and exporting American AI are the administration's immediate priorities. In the coming months, at least three questions will be of interest: 1) How does the administration navigate between a business-friendly approach while considering policy tools like compute export controls for competition with China? 2) How might wider cuts to scientific research funding affect the AI plan's strategic goals? 3) How will exporting the American tech stack intersect with other states' efforts to develop sovereign AI strategies? These will be important areas to watch.

Jessica Ji, Senior Research Analyst

The cybersecurity portions of the AI Action Plan establish a government-led AI Information Sharing and Analysis Center and direct work on AI-security-relevant toolkits and standards for the Department of Defense and Intelligence Community. Notably, the AI Action Plan delegates many security-related recommendations to the Center for AI Standards and Innovation (CAISI) or requests CAISI's involvement in recommendations. CAISI, a relatively new organization within the National Institute of Standards and Technology that facilitates testing and collaborative research to harness and secure AI systems, is already attempting to juggle a substantial list of priorities. CAISI may therefore struggle to implement all of the recommendations that have been assigned to them. I would also expect to see the Department of Homeland Security and the Cybersecurity and Infrastructure Security Agency continue to lead on certain initiatives, such as those related to critical infrastructure cybersecurity, due to their mandates and track record of leadership on critical infrastructure topics. Simultaneously, many of the Plan's recommendations build on existing cybersecurity infrastructure (such as vulnerability sharing mechanisms), which often relies on government funding. However, recent cuts to government capacity may impede efforts to maintain this foundational infrastructure.

Jacob Feldgoise, Senior Data Research Analyst

The AI Action Plan recommends new export controls on the sub-systems of semiconductor manufacturing equipment (SME). The proposed controls would aim to slow China's development of SME by denying the country critical foreign inputs. We have not yet seen the administration take action on this recommendation. Since October 2022, the U.S. government has imposed multiple rounds of export controls on advanced AI chips and SME, seeking to prevent China from using these items to develop advanced AI systems that could modernize its military. However, to date, China's own producers of SME have not been directly targeted by this regime; some controls on SME sub-systems are already in place, but they are not comprehensive. To design effective, new controls, the Department of Commerce will need to assess which sub-systems (1) would have the greatest delaying effect on SME development if denied to China, (2) currently have low foreign availability, and (3) would be difficult for China to develop and produce.

Ali Crawford, Senior Research Analyst

The AI Action Plan focuses on both empowering the American workforce and recognizing that the introduction of AI requires a reconfigured response through building AI literacy, strengthening workforce skills, and piloting innovative retraining programs while monitoring AI's impact on the labor market. The Plan builds on Executive Orders (EO) 14277 and 14278, extending federal efforts to align education and workforce development with AI adoption. EO 14277 established the Presidential AI Challenge, an ongoing initiative that encourages K-12 youth, educators, mentors, and community teams to solve real-world problems in their communities using AI. A guidebook published in August sets out goals, requirements, and prizes for the Challenge. Submissions to participate are due in January 2026, with the National Championship slated for June 2026. Regional champions will receive certificates of achievement, cloud credits, online resources, and a trip to Washington, D.C. National champions - across elementary, middle, high school, and educator divisions - will each be awarded $10,000. Essentially, the Challenge incentivizes communities to apply AI to real-world challenges, sharpening their AI skills and advancing the AI Action Plan's workforce priorities in the process.

For more details about the AI Action Plan, explore the Plan's themes and summaries in AGORA.

🏗️ Other AGORA news

• Devin Von Arx, one of CSET's 2025 summer interns, used AGORA data to publish an ETO blog and a companion CSET blog on California's approach to AI governance. In her work, she touches on SB 53, the transparency-focused California bill that requires developers to publish their safety protocols and implement whistleblower protections.
• The previous AGORA newsletter explored AI-related provisions in the House version of the One Big Beautiful Bill Act. You can now browse all AI-related provisions of the enacted bill in AGORA.
• The AGORA bulk dataset is being updated regularly with the latest data from the web UI. (Read the docs)

🤝 AGORA collaborations

• MIT AI Risk Initiative researchers Simon Mylius, Peter Slattery, Yan Zhu, Alexander Saeri, Jess Graham, Michael Noetel, and Neil Thompson teamed up with CSET's Mina Narayanan and Adrian Thinnyun to pilot an approach to map over 950 of AGORA's AI governance documents to several extensible taxonomies.
• The CSET AGORA team works closely with the Governance and Responsible AI Lab (GRAIL) at Purdue University to maintain and improve AGORA. Matthew Catani, a Research Assistant at GRAIL, also analyzed the AI Action Plan in the latest edition of the AI Policy Corner.

🏛️ CSET AI governance highlights

• CSET's AI Action Plan tracker monitors which agencies are responsible for implementing recommendations and the types of actions that agencies should take.
• Harmonizing AI Guidance: Distilling Voluntary Standards and Best Practices into a Unified Framework, a report from Kyle Crichton, Abhiram Reddy, Jessica Ji, Ali Crawford, Mia Hoffmann, Colin Shea-Blymyer, and John Bansemer, organizes over 7,000 recommended practices from 52 reports into a single unified framework.
• CSET Non-Resident Senior Fellow Pablo Chavez proposes a framework to guide U.S. international AI diplomacy in his report U.S. AI Statecraft: From Gulf Deals to an International Framework.

✉️ Get in touch

As always, we're glad to help you get the most out of AGORA and our other resources. Visit ETO's support hub to contact us, book live support with an ETO staff member, or access the latest documentation for our tools and data. 👋