SOCHI, June 10. /TASS/. The development of artificial intelligence technologies has significant implications for global biological security, as neural networks can potentially assist in the creation of dangerous biological weapons, head of the Russian sanitary watchdog Anna Popova stated at a plenary session of the 5th 'Global Threats to Biological Security. Problems and Solutions' international conference.
"The advancement of artificial intelligence also poses a threat to biological security," Popova stated. "Last January, the US company OpenAI revised its policy on the use of ChatGPT, lifting the ban on military applications. Subsequently, other developers followed suit."
She went on to cite alarming examples: "Last year, a group of biochemists in the US tested GPT’s capabilities. The program suggested methods for using pathogens in ballistic missiles to cause maximum damage. Also in 2023, the Massachusetts Institute of Technology conducted an experiment to assess how AI could assist non-experts in developing biological weapons. Within an hour, the system identified four pandemic-triggering pathogens and provided simple instructions on how to create them – along with guidance on how to avoid common mistakes."
Such developments warrant a reassessment of current approaches to biological security, Popova concluded.