{"id":15709,"date":"2026-08-06T21:16:30","date_gmt":"2026-08-06T21:16:30","guid":{"rendered":"https:\/\/nile1.com\/en\/?p=15709"},"modified":"2026-08-06T21:17:37","modified_gmt":"2026-08-06T21:17:37","slug":"stanford-researchers-use-generative-ai-to-create-first-synthetic-viruses","status":"publish","type":"post","link":"https:\/\/nile1.com\/en\/2026\/08\/06\/stanford-researchers-use-generative-ai-to-create-first-synthetic-viruses\/","title":{"rendered":"Stanford Researchers Use Generative AI to Create First Synthetic Viruses"},"content":{"rendered":"

Researchers at Stanford University<\/a> have used generative artificial intelligence to write complete biological genomes from scratch, successfully designing and creating functional synthetic viruses for the first time.<\/p>\n

The team utilized a genomic language model named Evo 2 to design 16 synthetic phages\u2014viruses engineered specifically to infect and destroy bacteria. In secure laboratory trials, the AI-generated viruses were tested against a non-pathogenic strain of the common microbe E. coli<\/a>. Researchers found that the synthetic phages performed more effectively at killing the bacteria than the naturally occurring phage \u03a6X174, which served as the operational template for the AI model.<\/p>\n<\/p>\n

\u201cOur study is a proof of concept showing for the first time that generative design can generate entire functional genomes,\u201d said project leader Brian Hie.<\/p>\n

To address biosecurity concerns, the Stanford researchers built specific safety boundaries into the software architecture. The underlying genomic language model was intentionally trained without exposure to viruses capable of infecting the complex cells found in plants and animals, confining its capabilities to bacterial targets.<\/p>\n

\"Column<\/p>\n

The advance comes as global healthcare systems face accelerating challenges from antimicrobial resistance, which has eroded the efficacy of standard antibiotic treatments. Phages have long been viewed as a prospective biological alternative to traditional antibiotics, though custom design has historically been limited by natural evolutionary constraints.<\/p>\n

\"Landscape<\/p>\n

According to another member of the Stanford team, AI could now help lead to a revival of phage therapy: \u201cWith more evidence coming out on the emergence of scary antibiotic-resistant pathogens, we will need innovative alternative solutions \u2014 and phage therapy is certainly one.\u201d<\/p>\n

\n

Read also:<\/h3>\n