Charting a Responsible Future in AI & Biosecurity: Understanding Technology Capabilities

The National Academies will host a webinar on November 6 on Charting a Responsible Future in AI & Biosecurity.

Organized by the Health and Medicine Division’s Forum on Microbial Threats and the Division of Earth and Life Studies’ Board on Life Sciences, this is the first of two sessions bringing together technology developers, researchers, and policymakers working at the intersection of foundation model-based artificial intelligence (AI) and biological research to explore the impact of these technologies on innovation, potential biosecurity risks, and promising solutions.

The first session on November 6 will examine the technological capabilities and bridge the needs for risk management between developers and policymakers.

Highlighted agenda items and strategic questions include:

• Overall framing: how close are we to achieving a full understanding of the genotype-to-phenotype relationship?
• Beneficial application of machine learning to develop tools that mitigate biosecurity risks
  and potential dual-use applications
• Application of foundation models to protein design and implications for biosecurity
• Draft framework on control points and addressing risks through each mechanism
• Considering the real vs projected risks landscape in life sciences research, and practical approaches to address each
• The true risks for misuse present day and in the future: What are the risks for intentional misuse? For accidental misuse?
• How do these new technologies lower the threshold for expertise in carrying out potential attacks?
• What are the necessary data points to inform the development of policy frameworks that balance the innovation and risk considerations?

The second session on November 15 will focus on governance of these emerging technologies, including challenges in defining regulatory policies as well as approaches to safeguard against these risks.