New viruses created with AI. For the first time, scientists have used artificial intelligence to design 16 new viruses intended to infect bacteria; they do not pose a threat to humans. The result was achieved by teams at Stanford University and the Arc Institute in Palo Alto, California. The objective, described in a study published in Science, is regarded as “a significant turning point” on the path toward developing new therapies. At the same time, the production of AI-designed viruses raises questions about safety and the risks associated with such work. “This is another step forward in the complexity that generative AI can design. For the first time AI is being used to design a complete genome, something that can replicate and perform other functions inside cells. It was uncharted territory for us,” says Brian Hie, an associate professor at Stanford University, to the BBC.

The researchers trained the AI to recognize patterns in DNA structure found in nature and to use that data to create “recipes” for producing entirely new viruses. They followed the AI’s instructions to produce DNA molecules that were then inserted into bacteria. The modified bacteria subsequently produced the novel viruses capable of infecting other bacteria. The scientists produced bacteriophages—viruses that infect only specific bacterial species. The 16 viruses proved effective at eliminating E. coli bacteria.

Professor Hie highlights that the new approach could “greatly improve human health” through the development of new drugs and therapies. However, there is the other side of the coin: concerns about misuse of AI. The BBC notes that Science includes a commentary by Professor Thomas Inglesby and Professor Moritz Hanke of the Johns Hopkins Center for Health Security: according to them, the results raise “urgent questions about biosafety and biocontainment.” The question is no longer whether generative design of viral genomes will exist, but whether it can be used without “causing serious harm.”