Safeguarding Against Global Catastrophe

Risks, Opportunities, and Governance Options at the Intersection of Artificial Intelligence and Biology

Overview

Imagine this: A global pandemic that results in more than 850 million cases, and 60 million deaths is sparked by a novel enterovirus strain that was intentionally engineered by an extremist group using artificial intelligence–enabled capabilities. This is the fictional scenario that senior leaders grappled with during a tabletop exercise hosted by NTI in partnership with the Munich Security Conference (MSC) in February 2025.

The exercise explored opportunities and risks at the convergence of AI and life sciences (AIxBio), and the resulting report offers practical recommendations to prevent catastrophic misuse.

The key takeaway? The exploitation of AIxBio capabilities for harm is a plausible near-term risk—and the time for action is now.

Exercise Findings

1. The fictional pandemic facilitated by the exploitation of AIxBio capabilities is a plausible, high-consequence bioterror scenario that technical experts and counter-terrorism professionals find deeply concerning and worthy of near-term action to prevent.
2. Rapidly advancing AIxBio capabilities are eroding barriers to bioweapons development by malicious actors, while also raising the ceiling of possible harms.
3. Existing security regimes are not equipped to address AI-enabled biological threats, but multiple intervention opportunities exist to reduce the risk that these capabilities could be exploited for biological weapons development.
4. AIxBio model developers can lead in developing best practices and technical solutions to reduce risks throughout the AI-model development life cycle.
5. There is a need for an official forum at the international level for communication and coordination to assess and address AIxBio risks.
6. Efforts to manage AI risks must take into account the benefits of AIxBio capabilities and avoid placing undue constraints on scientific benefits.

Recommendations

To address these findings, the report authors issue four recommendations:

1. National governments, industry, and philanthropic organizations should politically support and fund a comprehensive research and development agenda to establish effective guardrails for AIxBio technologies and applications.
2. National governments should establish institutes focused on AI safety and security, whose missions include safeguarding AIxBio capabilities.
3. International partners should collaborate via a global platform to develop and disseminate best practices for AIxBio safeguards.
4. Global scientific and national security communities should actively leverage AI’s beneficial applications in biosecurity and pandemic preparedness.

Why It Matters

AIxBio capabilities hold immense promise—but they also introduce unprecedented risks. Without decisive action, the race for AI supremacy could devolve into a race to the bottom on safety. National governments, industry leaders, and philanthropic organizations must champion AIxBio innovation where safety and security are not afterthoughts, but core drivers of progress—essential to unlocking the full positive potential of these powerful technologies.