Strengthening bio defense

Bioweapons pose an immediate danger. According to US intelligence, there are ongoing state programs dedicated to the development of bioweapons, despite the provisions in the Biological Weapons Convention intended to halt this work. There have also been multiple bioattacks by non-state actors within recent memory, both successful and unsuccessful.

To make things worse, the barriers to bioweapons development are falling rapidly. The cost of nucleic acid synthesis is dropping, and as AI advances, the opportunity to use new technologies for guidance throughout the bioweapons development process becomes increasingly viable, potentially lowering the bar for the expertise required to produce credible threats. Whether AI will meaningfully aid novices in creating biological weapons is hotly debated, and the measured effect of AI in this process was not significant as of mid-2025. But frontier models have continued to improve on every biology benchmark, including those measuring the tacit knowledge required for benchwork troubleshooting. This means the cost of bioweapons development is decreasing while the number of people who could successfully execute an attack is likely increasing.

Most concerning, as machine learning models for biology improve, bioengineering that bypasses normal evolutionary restrictions on the combined transmissibility and lethality of a virus may become more accessible. As the frontier moves forward, so too does the potential for biothreats we have never been confronted with before.

Biosecurity is uniquely difficult because bioattacks scale asymmetrically against practical defenses. There is a massive offensive advantage: compared with defenses such as pathogen-agnostic detection and medical countermeasures, attacks take less time to deploy, cost less, and benefit more from advances in AI and biotechnology. Broadly, attacks are relatively more bottlenecked by information and tacit knowledge, while defenses are relatively more bottlenecked by physical properties, such as manufacturing capacity and physical infrastructure, and coordination challenges. By removing friction in accessing information, including tacit knowledge, AI is poised to accelerate attackers’ progress more than defenders’.

There is one area where defense has an advantage in biowarfare: resources. Defenders have more access to capital, research infrastructure, and capable personnel. The US should use this advantage, including the likely influx of $37–100B/year in new AI-wealth-driven giving, to develop capabilities to prevent, detect, withstand, and combat engineered bioweapons, with the goal of ultimately deterring their development and deployment.