The artificial intelligence models underpinning chatbots could help plan an attack with a biological weapon, according to research by a US thinktank. A report by the Rand Corporation released on Monday tested several large language models (LLMs) and found they could supply guidance that "could assist in the planning and execution of a biological attack". However, the preliminary findings also showed that the LLMs did not generate explicit biological instructions for creating weapons. The report said previous attempts to weaponise biological agents, such as an attempt by the Japanese Aum Shinrikyo cult to use botulinum toxin in the 1990s, had failed because of a lack of understanding of the bacterium. AI could "swiftly bridge such knowledge gaps", the report said.

As always for a presidential inaugural, security and surveillance were extremely tight in Washington, DC, last January. But as George W. Bush prepared to take the oath of office, security planners installed an extra layer of protection: a prototype software system to detect a biological attack. The U.S. Department of Defense, together with regional health and emergency-planning agencies, distributed a special patient-query sheet to military clinics, civilian hospitals and even aid stations along the parade route and at the inaugural balls. Software quickly analyzed complaints of seven key symptoms-from rashes to sore throats-for patterns that might indicate the early stages of a bio-attack. There was a brief scare: the system noticed a surge in flulike symptoms at military clinics.