Inside the White House's AI Strategy

Kratsios says his goal in government is to help American technology lead by creating conditions for innovation. Policy decisions involve many agencies, so the White House brings them together; commercial drone rules, for example, must account for both FAA requirements and security around nuclear sites.

Kratsios says the administration supports both open and closed AI ecosystems, and points to the AI Action Plan’s commitment to open source. He argues that fixed regulatory thresholds can age poorly, citing the EU AI Act being finalized before ChatGPT and the difficulty of updating rules once they are set.

Kratsios says the White House seeks input from companies of different sizes and argues that a national AI standard could help startups avoid a patchwork of state rules. He distinguishes “born-free” technologies, such as the early internet, from “born-in-captivity” technologies such as commercial drones, where government approval can block commercialization.

Kratsios describes a new science report as a response to a changed research landscape: private companies and philanthropy now fund about 70% of U.S. R&D, while government funds about 30%. He expects AI to change scientific work and describes autonomous cloud labs that could run experiments, assess results, and test new hypotheses.

Kratsios says quantum computing deserves more attention and notes a federal goal to build a scientifically relevant quantum computer by 2028. He favors congressional action on a national AI framework and clearer protections around AI outputs and creators’ rights, while cautioning against legislating too quickly in areas where markets are still developing.