Trump's AI Shift: US AI Safety Institute's New Course

The departure of Elizabeth Kelly, the current director of the U.S. AI Safety Institute (US AISI), coincides with the Trump administration's shift in AI policy, as outlined in the Executive Order "Removing Barriers to American Leadership in Artificial Intelligence." This new direction has significant implications for the future of the US AISI and its mission to ensure the safety and responsible development of advanced AI systems.



The Trump EO, issued on January 23, 2025, replaces President Biden's Executive Order 14110, which emphasized oversight, risk mitigation, and equity in AI development. The new order prioritizes deregulation and the promotion of AI innovation as a means of maintaining US global dominance. This shift in policy has several potential implications for the US AISI's future funding, priorities, and partnerships.



Firstly, the Trump EO directs the White House to revise and reissue OMB AI memoranda to departments and agencies on the Federal Government's acquisition and governance of AI. This could potentially lead to changes in funding allocation for AI-related initiatives, including those led by US AISI. The Biden EO had directed federal agencies to collaborate in the development of best practices for AI safety and reliability, which may have contributed to US AISI's funding and resources. The Trump EO effectively halts these efforts, which could impact US AISI's budget and support.

Secondly, the Trump EO's emphasis on deregulation and AI innovation may influence US AISI's future focus and initiatives. The Biden EO had mandated extensive interagency cooperation to assess AI risks to national security and public welfare. The Trump EO's focus on deregulation and innovation could result in a reduced emphasis on risk assessment and mitigation by US AISI.

Lastly, the Trump EO's more unilateral stance on global AI leadership could impact US AISI's international partnerships and collaborations. The Biden EO had emphasized international cooperation, encouraging US engagement with allies and global organizations to establish common AI safety standards and ethical frameworks. The Trump EO's shift in global AI leadership approach could affect US AISI's international partnerships, such as its recent announcement of cooperation with Anthropic and OpenAI on AI safety research.

In conclusion, the Trump administration's shift in AI policy has potential implications for the U.S. AI Safety Institute's future funding, priorities, and partnerships. As the new leadership adapts to the Trump administration's focus on deregulation and AI innovation, they may need to reframe their mission and messaging, collaborate more closely with industry, and adapt their international engagement strategy. By doing so, the US AISI can continue to play a crucial role in shaping the future of AI in the United States while aligning with the administration's priorities.