The AI Policy Podcast: Unpacking the White House AI Action Plan with OSTP Director Michael Kratsios

Hosted by Gregory C. Allen and Michael Kratsios, July 31, 2025

1. Introduction

In this episode of The AI Policy Podcast, Gregory C. Allen engages in an in-depth conversation with Michael Kratsios, Director of the White House Office of Science and Technology Policy (OSTP). The discussion centers around the Trump administration's newly released AI Action Plan and its implications for securing U.S. leadership in artificial intelligence (AI). The episode, recorded during a CSIS-hosted event on July 30th, delves into the strategic policies shaping the future of AI in national security, innovation, and global competitiveness.

2. Background and Context

Gregory Allen sets the stage by reflecting on his long-standing collaboration with Director Kratsios, highlighting Kratsios's exemplary performance in public service. He references former Secretary of Defense Ash Carter’s insights on political appointees, praising Kratsios for being in the top tier of effective leaders from the outset of their collaboration (00:45).

3. Evolution of U.S. AI Policy

Ole Kratsios and Allen trace the origins of the current AI policy trajectory back to the first Trump administration. They emphasize the significance of President Trump’s February 2019 executive order, which marked the first national AI strategy by the U.S., launching the American AI Initiative and doubling AI-related spending. This foundational step positioned the U.S. to lead in AI research and development long before the advent of technologies like ChatGPT (02:12).

Allen critiques the subsequent Biden administration’s approach, characterized by a more restrictive AI executive order that imposed stringent pre-deployment testing and fostered apprehension within the AI community. He argues that fear-driven policies hindered innovation and led to unnecessary regulatory burdens (04:30).

4. The Trump Administration’s AI Action Plan

Upon re-entering office, President Trump rescinded the Biden executive order and swiftly introduced a comprehensive AI Action Plan. The plan is structured around three primary pillars:

1. Innovation: Maintaining U.S. leadership in AI by fostering a robust research and development ecosystem. This involves creating a regulatory environment conducive to innovation, providing certainty to AI companies, and investing in critical R&D for technological advancements (06:00).

2. Infrastructure: Building the necessary physical infrastructure, including data centers and power generation capabilities, to support the burgeoning AI industry. The plan addresses bottlenecks in electrical generation and data center construction, aiming to streamline permitting processes and accelerate development (06:45).

3. Exports: Promoting the American AI stack globally to ensure international reliance on U.S. AI technologies. This pillar focuses on offering comprehensive packages that include chips, models, and applications, making it easier for other countries to adopt American AI solutions (07:30).

5. AI Exports in Detail

Kratsios elaborates on the export strategy by drawing parallels to the Huawei situation. Unlike Huawei’s fragmented stack, the American approach offers turnkey solutions that simplify the adoption process for international clients. This strategy aims to make American AI technologies the default choice worldwide by providing bundled packages that cover the entire technology stack (10:59).

Brielle Hill adds that the executive order leverages U.S. development finance organizations, such as the Export-Import Bank and the Development Finance Corporation, to prioritize AI exports. These institutions will create attractive financing packages to support the global dissemination of American AI technologies, ensuring that U.S. standards and innovations become the global norm (15:05).

6. Export Controls and National Security

The action plan balances promoting AI exports with strengthening export controls to safeguard national security. Hill emphasizes the importance of controlling the export of high-end semiconductors to adversarial nations like China, ensuring that critical technologies do not fall into the wrong hands. This involves enhancing the enforcement capabilities of the Bureau of Industry and Security (BIS) to prevent illegal exports and smuggling of advanced chips (17:47).

Allen commends the administration for integrating intelligence community efforts into export control enforcement, a move he describes as long overdue. This interagency collaboration aims to tighten the grip on unauthorized exports and maintain U.S. technological supremacy (19:01).

7. Regulation and Governance

The discussion shifts to AI regulation, contrasting the U.S. approach with the European Union’s AI Act. Hill criticizes the EU’s broad, precautionary regulatory framework, arguing that it stifles innovation by imposing uniform rules across all sectors. In contrast, the U.S. strategy advocates for sector-specific, risk-based regulations managed by industry experts within respective regulatory agencies (e.g., FDA for medical AI applications). This nuanced approach allows for more adaptable and effective oversight without hindering technological progress (25:40).

8. State-Level AI Regulation and Preemption

Addressing the potential fragmentation of AI regulations at the state level, Hill highlights the administration’s efforts to advocate for uniform federal regulations. The goal is to prevent a patchwork of state laws that could burden startups and small businesses with conflicting compliance requirements. While some executive measures are being explored, Hill acknowledges that comprehensive preemption would require bipartisan legislative action from Congress (28:37).

9. AI and Labor

Hill underscores the critical need to prepare the American workforce for the AI-driven economy. The AI Action Plan includes initiatives to reskill and retrain workers to support AI infrastructure projects, such as data center construction. She cites the example of the Stargate project in Texas, which faces a shortage of electricians, illustrating the broader labor challenges in scaling AI infrastructure (38:15).

Additionally, the administration has committed to K-12 AI education, ensuring that students are equipped with the knowledge and skills to thrive in an AI-integrated job market. This proactive approach aims to secure high-paying, future-proof jobs for Americans across various industries (40:50).

10. AI and CBRNE/Biosecurity Risks

While acknowledging the importance of addressing AI-related risks, Hill argues that the administration seeks to integrate risk evaluation into a broader AI strategy rather than letting it dominate policy decisions. The focus is on managing specific risks, such as those related to chemical, biological, nuclear, radiological, and explosive (CBRNE) threats, without hindering overall AI progress. This balanced approach ensures that safety considerations are addressed without compromising innovation (32:09).

11. Center for AI Standards and Innovation

The conversation touches on the future of the Center for AI Standards and Innovation (formerly CASEY AI Safety Institute). Hill explains that the center will transition from a safety-focused institute to a standards body, aligning with NIST’s mission to develop measurement standards for AI models. This shift aims to provide the AI industry with standardized evaluation methodologies, enhancing trust and reliability in AI applications across sectors (35:28).

12. Legislative Priorities and Future Steps

While the AI Action Plan leverages executive authorities to enact significant changes, Hill acknowledges areas where Congressional support is essential. Key legislative priorities include:

• Preemption of State Regulations: Ensuring uniform AI regulations across all states to prevent regulatory fragmentation.

• Fair Use and Copyright: Addressing legal ambiguities surrounding the training of large language models on copyrighted materials.

• Statutory Support for Standards Institute: Providing legal frameworks to empower the Center for AI Standards and Innovation to develop and implement standards.

• R&D Funding: Prioritizing AI-related research and development funding across agencies like NSF to sustain innovation (47:24).

13. Conclusion

In closing, Gregory C. Allen commends Michael Kratsios and the Trump administration for the comprehensive AI policies outlined in the AI Action Plan. He emphasizes the plan’s potential to position the United States as a global leader in AI, fostering innovation while safeguarding national interests. Hill reiterates the administration’s commitment to smart, balanced regulation that promotes growth and addresses critical risks without stifling technological advancements.

Notable Quotes:

• Gregory C. Allen (00:45): "I knew from the first meeting I had with Director Kratzios that he was in that best 25% who really nail it."

• Brielle Hill (06:00): "America has to have the most dominant technological stack in the world. That's critically important for our national economic security."

• Gregory C. Allen (19:01): "Using the White House to actually drive that kind of connection between those various agencies, I think that that's so long overdue."

• Brielle Hill (25:40): "The key fundamental distinction between the way that the Europeans view AI regulation and we do as administration is the Europeans continue to think of it as a singular AI regulation that is sort of horizontal and crosses everything."

• Brielle Hill (38:15): "We have to start building the infrastructure necessary to power the AI economy, and that means reskilling and retraining our workforce."

This episode provides a comprehensive overview of the Trump administration's strategic approach to AI policy, emphasizing innovation, infrastructure, and global competitiveness. Through insightful dialogue, Allen and Kratsios illuminate the nuanced balance between fostering technological advancement and ensuring national security, offering valuable perspectives for policymakers, industry leaders, and the AI community at large.

Timestamps:

• 00:45: Introduction by Gregory C. Allen
• 02:12: Background on Trump administration's AI initiatives
• 06:00: Overview of the AI Action Plan's three pillars
• 06:45: Detailed discussion on Infrastructure
• 07:30: Insights on AI Exports
• 10:59: Export strategy and comparison to Huawei
• 15:05: Leveraging development finance organizations for AI exports
• 17:47: Balancing AI exports with export controls
• 19:01: Interagency cooperation on export control enforcement
• 25:40: U.S. vs. EU AI regulatory approaches
• 28:37: State-level AI regulation and preemption efforts
• 32:09: AI and CBRNE/Biosecurity risk management
• 35:28: Future of the Center for AI Standards and Innovation
• 38:15: AI and labor: training and education initiatives
• 40:50: K-12 AI education executive order
• 47:24: Legislative priorities and Congressional involvement

For more information, visit csis.org.