Improving Biosecurity with A Three-legged Stool Approach

A pandemic is not a once in a century event, and the international community must prepare now for the next one. This includes recognizing that the next global biological crisis could be the result of an accidental release or a deliberate attack. Protecting against these types of biological threats requires novel approaches to develop effective safeguards and oversight mechanisms, especially given the rapid advancement of emerging technologies and the potential for these powerful tools to be misused for harm.

The use of AI makes it easier to manipulate biology to achieve a desired end-state in a shorter amount of time as compared to the traditional experimental process. For example, AI-enabled capabilities can allow scientists to develop targeted medical countermeasures to more effectively prevent or treat disease. These tools are already being applied in pandemic preparedness to help researchers predict how a virus will evolve to create more protective vaccines. However, a malicious actor could abuse these same capabilities, using them to alter a pathogen to evade available countermeasures, or to make pathogens more virulent or more transmissible among people. Without guardrails, AI-enabled tools could make it easier to deliberately misuse biological knowledge, tools, and techniques, and the consequences could be catastrophic.

“We are experiencing a 21st century revolution in bioscience and biotechnology… [with] incredible potential benefits to society for advancing human health, combating the effects of climate change, and furthering economic development …,” explained Jaime Yassif, Vice President, Global Biological Policy and Programs at NTI (NTI | bio). “But at the same time a lot of these advances are dual use, and this poses a risk of accidental or deliberate misuse that could have catastrophic consequences.”

At the 2025 Munich Security Conference (MSC), Dr. Yassif, joined other senior leaders to discuss the potential benefits and risks posed by the convergence of AI and the life sciences. The discussion, a panel during the Emerging Threats Forum called Protecting the World from Pathogens: Biosecurity in the Age of Transformational Tech focused on ways to reduce gaps in biosecurity to pave the way to a safe, innovative future.

The panel included:

• Dr. Andrew Hebler, Director of Biosecurity at the Coalition for Epidemic Preparedness Innovations

• Dr. Tom Inglesby, Director Center for Health Security at Johns Hopkins University

• Dr. Jaime Yassif, Vice President for Global Biological Policy and Programs at the Nuclear Threat Initiative

• Moderator: Dr. Mayesha Alam, Senior Vice President of Research at Foreign Policy Analytics

To address this challenge, Dr. Yassif described a three-legged stool approach to improve biosecurity and governance of rapidly advancing AIxBio capabilities:

1. Develop practical, technical solutions to reduce risks: Built-in safeguards for AI-enabled biodesign tools can decrease the risk of misuse, as an example.
2. Leverage global platforms to share best practices: Collaboration among national policymakers, expert communities, and the private sector can foster sharing and shaping of best practices so they can be put into practice internationally
3. Engage governments and private funders to incentivize compliance: Decision makers should deploy measures to incentivize compliance with practical biosafety and biosecurity best practices that guard against misuse of AIxBio capabilities while also supporting innovation and beneficial applications.

Further, the panelists all agreed that there is significant interest within the biosecurity community to balance benefits and risks and engage in dialogue about safe and responsible innovation.

“The greatest risks are the potential that AI systems could catalyze, accelerate, or enable new pandemic risks, whether its resuscitation of old viruses that created pandemics or the creation of new ones,” noted Dr. Inglesby.  He emphasized that it is in all our interest to limit these systems, so they do not “…create a great harm to humanity.”

It is essential the international community strike a balance between fostering and supporting beneficial AIxBio innovation while guarding against accidental or deliberate misuse of these tools. NTI is working to address these challenges and offer solutions.

• NTI | bio hosted a tabletop exercise at the Munich Security Conference to create better understanding of the risks we face at the intersection of AI and the life sciences and identify solutions to reduce them.
• NTI | bio also published a report on Developing Guardrails for AI Biodesign Tools which describes how built-in guardrails and managed access can safeguard against misuse.
• The AIxBio Global Forum organized by NTI | bio brings together high-level biosecurity professionals, AI experts, and policymakers to promote the adoption of national and global governance mechanisms that support stronger biosecurity for AIxBio capabilities.
• Signatories of the Bio Funders Compact, a joint initiative by NTI | bio and CEPI, commit to incorporating biosecurity and biosafety into their decision-making processes to incentivize best practices and prevent unduly risky research from moving forward.

The Emerging Threats Forum, hosted by Foreign Policy, was an official side event of the Munich Security Conference presented in partnership with the Coalition for Epidemic Preparedness Innovations (CEPI), Gates Foundation, General Motors, Microsoft, NTI | Bio, Sentinel Bio, and the Special Competitive Studies Project (SCSP).