Stanford researchers use generative AI to design functional viruses

generative AI synthetic biology bacteriophages E. coli biosecurity

Researchers from Stanford University, the Arc Institute, and the Broad Institute have successfully used generative AI to design and synthesize complete, functional viral genomes from scratch. The study, published in the journal Science, utilized genomic language models named 'Evo 1' and 'Evo 2' to predict genetic sequences by learning from millions of existing DNA patterns.

The AI-designed organisms are bacteriophages—viruses that specifically infect bacteria and do not pose a threat to humans or animals. From approximately 300 chemically synthesized designs, 16 viable bacteriophages were produced and tested in a laboratory. These artificial viruses were effective at targeting and destroying E. coli bacteria, with some performing even better than naturally occurring phages and successfully overcoming existing bacterial resistance.

While the technology offers significant medical promise, particularly for developing new phage therapies to combat antibiotic-resistant "superbugs," it has also raised urgent biosecurity concerns. Experts warn that the ability to compose functional viral genomes via AI could be misused to create harmful pathogens, highlighting a gap in current regulatory and governance frameworks for managing rapidly advancing generative biology tools.

Entities

Arc Institute · Brian Hie · Broad Institute · E. coli · Escherichia coli · Evo 1 · Evo 2 · Evo1 · Evo2 · Harvard University · MIT · Science

Claims

What the coverage asserts, and how many sources carry each claim.

[● 25 SOURCES] The 16 functional viruses were designed to infect and destroy E. coli bacteria. www.digitalnewsreport.com · massinformacion.com.mx · www.nosabesnada.com · quo.eldiario.es · www.osul.com.br · +20 more
[● 23 SOURCES] The research utilized AI models known as Evo1 and Evo2. www.diariodemallorca.es · massinformacion.com.mx · www.nosabesnada.com · www.osul.com.br · www.timesnownews.com · +18 more
[● 22 SOURCES] Current governance and regulatory frameworks may be insufficient to manage rapidly developing generative biology tools. expreso.press · www.naftemporiki.gr · www.digitalnewsreport.com · www.epochtimes.com · www.timesnownews.com · +17 more
[● 26 SOURCES] Out of nearly 300 chemically synthesized AI-designed genomes, 16 resulted in viable bacteriophages. www.digitalnewsreport.com · www.diariodemallorca.es · massinformacion.com.mx · www.nosabesnada.com · quo.eldiario.es · +21 more
[● 24 SOURCES] The AI-designed phages were effective at overcoming bacterial resistance in E. coli. www.digitalnewsreport.com · www.diariodemallorca.es · www.ygeiamou.gr · massinformacion.com.mx · baoquocte.vn · +19 more
[● 24 SOURCES] Experts warn that the ability to compose viral genomes via generative AI raises urgent biosecurity and bioprotection concerns. expreso.press · www.naftemporiki.gr · www.digitalnewsreport.com · massinformacion.com.mx · baoquocte.vn · +19 more
[● 17 SOURCES] The bacteriophages created do not pose a threat to humans or animals. www.digitalnewsreport.com · massinformacion.com.mx · www.nosabesnada.com · quo.eldiario.es · www.osul.com.br · +12 more
[● 27 SOURCES] Researchers at Stanford University, the Arc Institute, and the Broad Institute used generative AI to design complete, functional viral genomes from scratch. www.digitalnewsreport.com · www.diariodemallorca.es · massinformacion.com.mx · www.nosabesnada.com · quo.eldiario.es · +22 more
[● 2 SOURCES] Approximately 300 genomes were chemically synthesized and laboratory tested. www.albinfo.ch · infa.lt
[● 2 SOURCES] The study utilized genomic language models named 'Evo 1' and 'Evo 2'. www.albinfo.ch · infa.lt
[● 2 SOURCES] The process resulted in the creation of 16 viable artificial bacteriophages. www.albinfo.ch · infa.lt
[○ 1 SOURCE] The technology raises concerns regarding dual-use risks in biological research. www.dcomedieta.it