What impact will advancements in artificial intelligence (AI) have on national security and global public health issues, particularly when it comes to the new AI capabilities that can facilitate the creation of novel infectious agents and new infectious disease outbreaks?
On May 28, 2026, Northwestern University’s Network for Collaborative Intelligence (NNCI) held a panel discussion in Chicago to assess the intersection of AI, biomedical research, biosecurity, and ethics. (View the Northwestern University panel recording.) The panel included this article’s authors, Daniel Eisenman and Mohammad Hosseini, who are both members of the Advarra AI Council:
| Name | Organization | Expertise |
| Mohammad Hosseini, MA, PhD | Northwestern University Feinberg School of Medicine | AI ethics, research integrity, and open science |
| Daniel Eisenman, PhD, RBP, SM(NRCM), CBSP | Advarra | Biosafety, biosecurity, AI, and emerging technologies |
| Doni Bloomfield, JD | Fordham University School of Law | Intellectual property, biosecurity, national security law, and health law |
| Nicole Kikendall, PhD | Argonne National Laboratory | Biosafety, biosecurity, and national security |
The panel discussed how AI and robotics are changing healthcare and biomedical research by expanding data analysis and predictive capabilities, enhancing drug discovery, and enabling new approaches to high-throughput diagnostics. In 2024, the Nobel Prizes in physics and chemistry recognized advances in artificial neural networks and protein structure prediction, which the panelists viewed as key developments in machine learning with significant implications for the future of medicine. At the same time, the panelists noted that the capabilities that make these advances possible may also introduce serious risks when used maliciously or carelessly, including the potential creation of novel toxins or infectious agents.
Another point of discussion was the growing accessibility of biological AI tools. Recent work involving AI-designed viruses illustrates both the potential of these tools to support therapeutic innovation and the risks associated with misuse or unintended harm. As open-source AI models and public datasets become more widely available, barriers to accessing sophisticated biological design capabilities are likely to continue to decrease. The panelists emphasized that biosecurity regulations may need to expand beyond controls for physical materials, such as biological agents and laboratory equipment. Future approaches may also need to address digital assets and workflows, including digital infrastructure, data access, and information flows.
The discussion also highlighted the limits of policies tied only to national jurisdictions or individual funding sources. Without international policies and guidelines, individuals or organizations may operate in jurisdictions with fewer oversight mechanisms. Similarly, requirements established by research funders—including the National Institutes of Health (NIH) and the National Science Foundation (NSF)—generally apply only to projects supported by those institutions. They may not address activities conducted through other funding sources.
The discussion also pointed to the need for coordinated governance across institutions and jurisdictions. Recent developments in this direction include the May 2025 White House Executive Order titled “Improving the Safety and Security of Biological Research” and the 2026 industry open letter “In Support of Mandatory Nucleic Acid Synthesis Screening and Recordkeeping.” Both have called for safeguards governing the creation of synthetic genetic material. However, the panelists noted several challengesobstructing the creation of effective AI regulation:
- AI risks are global concerns, and policies from any one country are unlikely to mitigate global threats on their own.
- Since the U.S. is a leading developer of AI and hosts substantial supporting infrastructure, U.S. policymakers may be reluctant to restrict AI development amid competition with China.
- AI concerns aren’t monolithic. Effective regulation and oversight will need to address a range of complex, interconnected issues.
- AI is developing quickly, requiring agile regulatory approaches and systematic efforts to anticipate high-risk implications.
- Policymakers need input from stakeholders with diverse technical expertise.
The panelists agreed that no single regulation, intervention, or safeguard will sufficiently address the full range of AI-related biosecurity risks.
Managing these challenges will require a coordinated approach combining data governance, institutional oversight, technical safeguards, and expertise across AI, biosafety, biosecurity, and cybersecurity. As AI capabilities continue to evolve, ongoing collaboration among researchers, technology developers, institutions, policymakers, and security experts will help support responsible innovation and address emerging risks.
