Emerging advances in artificial intelligence (AI) present both opportunities and risks in the fight against infectious diseases, underscoring a need for comprehensive national biodefense strategies, experts say. César de la Fuente, director of the Machine Biology Group at the University of Pennsylvania, emphasizes that while AI holds promise for accelerating medical countermeasures such as new antibiotics, it may also lower barriers for the misuse of biological knowledge.

De la Fuente warns that AI’s ability to assist users beyond traditional scientific literature—for example, by explaining complex concepts and troubleshooting experimental failures—could enable individuals without formal expertise to navigate and exploit biological information for harmful purposes. This concern arises amid rapidly increasing public access to powerful digital tools, contrasting starkly with the slower pace of developing the laboratory infrastructure and public health systems needed to detect and respond to biological threats.

Recent developments highlight the urgency of updating biodefense efforts. Earlier this month, the World Health Organization investigated the unexplained death of a worker at a plague research institute in Russia, raising questions about oversight and preparedness. Meanwhile, a 2025 report from Britain’s AI Security Institute indicated that nonexperts using advanced AI models had a significantly higher likelihood of generating viable experimental plans in virology than those relying on standard internet searches. Similarly, the 2026 International AI Safety Report noted that major AI developers have introduced safeguards after recognizing their technologies might inadvertently aid the development of biological weapons by novices.

Despite these concerns, producing biological weapons still requires access to physical materials and specialized practical skills, factors not addressed by AI outputs alone. Nevertheless, de la Fuente argues that uncertainty regarding AI’s potential misuse should not delay proactive preparedness measures, which require years to implement.

The United States has existing frameworks, such as the National Biodefense Strategy updated in 2022 and the 2025 White House AI Action Plan, which call for enhanced screening of synthetic genetic materials and assessments of national security risks tied to AI. Experts recommend integrating these efforts into a coordinated, well-funded program that rigorously evaluates the country’s readiness to counter AI-enabled biological threats. This would involve designating a senior official accountable for overseeing execution and establishing public milestones.

De la Fuente calls for collaboration among technology companies, academic institutions, public health agencies, and national security bodies to conduct comprehensive risk assessments. He advocates for independent, qualified evaluators to conduct secure testing of AI models with potentially dangerous biological capabilities prior to release and following major updates. These evaluations would measure the degree of assistance AI provides beyond existing tools and the effectiveness of built-in safeguards. Developers would be required to report significant safety failures to authorities, with findings transparently shared with the public to the extent that security is not compromised.

Universities are urged to accept independent scrutiny of sensitive biological research while balancing openness with risk management. De la Fuente notes that restrictions should be evidence-based and flexible, emphasizing the need to support smaller laboratories to prevent security protocols from concentrating research in well-funded institutions alone.

Beyond preventive measures, de la Fuente stresses the importance of robust response capabilities. Regular nationwide exercises simulating AI-assisted biological threats should involve researchers, healthcare providers, manufacturers, and local health officials to test coordination and timing of interventions. Sustained investments are necessary to maintain laboratory capacity, trained personnel, and public health infrastructure capable of recognizing and managing unusual outbreaks or attacks.

Public involvement is also recommended in decisions regarding health data use and resource allocation, including transparency about preparedness exercises and corrective actions. On the international stage, the United States should seek cooperation through common safeguards and information sharing to enhance collective biodefense, recognizing that global preparedness reinforces domestic security.

De la Fuente concludes that while AI offers transformative potential for combating infectious diseases, urgent, deliberate action is essential to mitigate the risks posed by its misuse and protect public health.