The following statement can be attributed to Paul Lekas, Senior Vice President, Global Public Policy, the Software & Information Industry Association.
We welcome BIS’ announcement to replace the AI Diffusion Rule, proposed in January 2025, with a simpler version. Though the Rule recognized the impact of AI diffusion on U.S. national security, foreign policy, and economic interests, we believe it would ultimately undermine U.S. leadership. Its complex system of export licenses for AI model weights and compute would have created serious compliance and practical challenges and the proposed tiering of countries would have opened the floodgates for the PRC. The new rule should take a risk-based approach to transfers of AI models and compute, recognizing, as OSTP Director Michael Kratsios recently said, “it is not enough to seek to protect America’s technological lead. We also have a duty to promote American technological leadership.” The Rule did not grapple sufficiently with the value of adopting U.S. technology as a key component of U.S. national security and foreign policy. We anticipate the new rule will feature a risk-based approach to export controls focused on “chokepoint” technologies that also advances Kratsios’ vision of making the United States “the global partner of choice and the standards setter to follow if we enable and encourage American companies to distribute the American tech stack around the world.”
