A consortium of leading global security and technology experts has issued a stark warning: the rapid advancement of artificial intelligence is significantly lowering the technical barriers for creating engineered biological weapons, posing a near-term risk of a deliberate pandemic.
The warning stems from a tabletop exercise convened in February by the Nuclear Threat Initiative (NTI) and the Munich Security Conference (MSC). Fourteen experts from fields including biotechnology, AI, and counter-terrorism grappled with a hypothetical scenario in which a non-state actor uses AI to design a novel, highly contagious enterovirus. The simulated pandemic resulted in an estimated 850 million infections and 60 million deaths worldwide.
The group's consensus, detailed in a subsequent report, found this scenario "deeply concerning and worthy of near-term action to prevent." They concluded that existing and emerging AI tools could be misused to design pathogens with pandemic potential, and that current global biosecurity frameworks are inadequate to meet this evolving threat. The experts emphasized that the accelerating capabilities of AI are "eroding barriers to bioweapons development by malicious actors."
The report calls for urgent international cooperation to assess and govern these risks. It also stresses the need for a balanced approach that mitigates dangers without stifling the significant benefits AI promises for medical research and drug discovery.