Written Statement of Alexander J. Titus, PhD, Before the U.S. Senate AI Insight Forum on “Risk, Alignment, & Guarding Against Doomsday Scenarios”

<p>Artificial intelligence (AI) is poised to transform every domain of science and society, but in the life sciences, it carries both extraordinary promise and unprecedented peril. In testimony before the U.S. Senate AI Insight Forum on “Risk, Alignment, &amp; Guarding Against Doomsday Scenarios,” Dr. Alexander J. Titus argues for a refined, evidence-based approach to policy and governance of AI in biology. Drawing on his experience as Principal Scientist at the University of Southern California’s Information Sciences Institute and as the Department of Defense’s inaugural Principal Director for Biotechnology, Titus emphasizes the need to quantify AI-enabled biosecurity risks through empirical baselines that measure how much AI increases risk beyond existing capabilities. He introduces the concept of <i>Violet Teaming</i>—a multidisciplinary expansion of red teaming that integrates economists, ethicists, providers, and social scientists—to balance the trade-offs between innovation and risk mitigation. The testimony warns against both over- and under-stepping: overregulation that could hinder lifesaving advances in medicine, food security, and sustainability, and under-regulation that could fail to anticipate novel biological threats. Titus concludes that responsible progress depends on three imperatives: context-specific action, diverse stakeholder input, and systematic measurement of risk over time. The same AI systems that raise new threats, he notes, can also power autonomous science for rapid defense and response, making responsible deployment not only a matter of safety, but of national and global resilience.</p>