A teachable moment for dual-use
Fabio Urbina,Filippa Lentzos,Cédric Invernizzi,Sean Ekins
DOI: https://doi.org/10.1038/s42256-022-00511-6
IF: 23.8
2022-07-13
Nature Machine Intelligence
Abstract:To the Editor — Dual-use research of concern (DURC) can be defined as research, mainly in the life sciences, that has the potential to be misapplied for harmful purposes 1 . Key examples are the synthesis of mousepox 2 , the synthesis of poliovirus 3 , the generation of the 1918 influenza virus 4 , gain-of-function studies with H5N1 in ferrets 5 and the synthesis of horsepox, the viral cousin of smallpox 6 . These seminal events involved the physical synthesis of a biological agent. However, the time may have come to also consider dual-use risk of the development of toxic agents in silico, in the light of the alarming results of a computational experiment we recently performed for a biennial arms control conference 7 ; we used a generative artificial intelligence (AI) approach previously developed for drug discovery applications, and found it could easily design a range of nerve agents including VX. The experiment demonstrated the alarming speed and ease with which such software — based on open-source tools and datasets from the public domain — could be used for bad purposes. Our experiment was subsequently covered widely in the media, reaching a network of scientists, experts and lay people alike 8 , and its implications were recognized at the highest levels of governments within a matter of days. The level of interest was probably amplified owing to the enfolding war in Ukraine, with Russia's invasion and the threat of biochemical weapons use.
computer science, artificial intelligence, interdisciplinary applications