AI bioweapon fears are hiding real biological risks, warns expert

Twenty-five years after the 2001 US anthrax attacks, excessive public focus on artificial intelligence biosecurity risks is distracting society from the urgent need to regulate the broader life-sciences sector. The Amerithrax attacks caused 22 cases of pulmonary anthrax and 5 deaths, exposing poor government coordination and widespread panic. The FBI later traced spores to a US Army lab, though primary suspect microbiologist Bruce Ivins died before facing trial. Today, cheaper and more accessible equipment like benchtop DNA synthesizers and cloud labs allow private facilities to bypass federal checks, while AI reduces the knowledge barrier needed to design bioweapons. However, converting digital AI designs into functional biological products still faces steep physical hurdles, according to RAND biologist Barbara Del Castello. Regulatory oversight remains sluggish and fragmented, worsened by shifting US federal rules in 2025 and 2026 that banned seven types of dangerous gain-of-function studies outright.

Context

Biological safety regulations have historically struggled to keep pace with rapid technological advances across both public and private research facilities. Past crises like the 2001 anthrax letters and the COVID-19 pandemic show that systemic gaps in biosurveillance and lab oversight recur despite billions spent on biodefence.