Scientists have used artificial intelligence to design complete, functioning viruses with genomes never seen before in nature, a breakthrough that could open new avenues for fighting drug-resistant bacterial infections.
A Stanford University-led research team used generative AI models trained on millions of genomes to design viruses, known as bacteriophages, that can kill bacteria like E. coli, according to a study published Friday in the journal Science.
Researchers chemically synthesised and tested nearly 300 of the most promising designs, ultimately producing 16 viable phages capable of infecting bacteria that had evolved resistance to natural bacteriophages, The Hill reported.
When E. coli developed natural resistance to the phage used as the template for the experiment, a mixture of the AI-designed phages was able to rapidly overcome that resistance, while a comparable mixture of naturally sourced phages was unable to do so, according to the research.
The study is a major step toward preventing deadly, unstoppable pathogens and is the first time researchers have demonstrated the emerging technology’s potential power in this space.
The AI-generated phages were not just copies of viruses already found in nature, according to the authors, who noted the process contained new mutations and divergent genes.
But the ability to use AI to design entire functional viral genomes could raise biosecurity concerns and bring additional complications to the federal government’s efforts to oversee high-risk biological research.
The Trump administration issued a policy in June to regulate high-risk life sciences research, including experiments that could make harmful biological agents more dangerous.
Experts have warned about the security implications of the technology, as AI could increasingly allow researchers to design entire biological systems rather than simply modifying ones that already exist.
