Scientists Create First Viruses Designed By AI, What Does This Mean For The Future Of Medicine?
Scientists have created the first viruses that were designed using artificial intelligence. This feat is being highly debated within the scientific community as both a major milestone advancement for medicine, and a potential threat if this technology isn’t protected.
Dr. Brian Hie is a chemical engineer at Stanford University in California, who utilized genome language models to design functioning genomes for bacteriophages. The models are the genetic equivalent to the language models used for AI chatbots, and bacteriophages are the specific viruses that only infect bacteria, so they’re used to treat patients with persistent infections.
The viruses were then made in the laboratory and specifically put against E. coli in a dish, according to the research recently published in the journal Science. “The ability to rapidly design genomes and tune them for specific bugs while overcoming resistance could transform phage therapy and expand biotechnological toolkits,” the researchers wrote.
On the other hand, the scientists have also stated that the new work raised “important biosafety, biocontainment, and biosecurity consideration,” telling other people who are designing whole genomes to “consult both safety and security professionals throughout the project.”
Professor Tom Inglesby and Dr. Mori Hanke at the Center for Health Security at Johns Hopkins University wrote an accompanying article in which they emphasized the researchers safety warnings:
“Although this is promising for life sciences applications, it also raises urgent biosafety and biosecurity questions. The ability to compose viral genomes using generative AI now exists; the governance to safely steer it does not.”
Dr. Hie and his colleagues mapped out in their article how they used AI models named Evo1 and Evo2 to design the new viral genomes. The models were specifically trained with genetic data from 2 million bacteriophages.
To keep the focus on the bacteriophages that can be utilized in medicine, the scientists made sure to carefully exclude any genetic codes for viruses that can infect living beings so the AI training doesn’t design any dangerous viruses.
The AI worked to generate thousands of potential genomes, and the researchers chose about 300 to then make in a lab. Those genomes were then dropped into bacteria which read the genetic code and created the new bacteriophages.
Only 16 bacteriophages were viable, however, the makeup of them were able to overcome resistance in two different strains of E. coli.
Professor Inglesby and Dr. Hanke said that although only 16 out of the 300 were viable, the work proves that generative AI can create viral genomes, and there should be safeguards put into place to ensure that the technology can never be used to create pathogens that can infect plants, animals, and especially humans.
“Such genomes might encode new pathogens that … cannot be contained by existing countermeasures,” they wrote.
Tom Ellis, a professor of synthetic genome engineering at Imperial College London, stated that this work was very impressive, but when it comes to creating more complex genomes, there’s a lot more work to actually be done.
“This is literally the smallest and easiest genome to make,” he said.
“An AI trained on the genetic code of dangerous bugs could be used to design more harmful viruses,” the Guardian reported that Ellis said.
“But controlling access to genetic data and having restrictions on making genomes that look dangerous would help. Governments are working hard to do this already,” he added.
“But honestly, the threat from full AI design and writing of a genome of a virus or bacteria is very overblown when we consider that just taking existing pathogens and making gain-of-function changes to their genomes is so much easier and much more likely to be a real pathogenic threat.”
Dr. Filippa Lentzos, a reader in science and international security at King’s College London, stated that one of the biggest things to now look out for with this type of work is making sure no scientific groups try to manufacture anything involving DNA.
“It’s important to see the bigger governance picture and not focus regulation solely on the AI model,” she said.
“A layered approach makes more sense: safeguards around model development and access, responsible research review, synthesis screening, and established laboratory biosafety and biosecurity.”
Eric Mastrota is a Contributing Editor at The National Digest based in New York. A graduate of SUNY New Paltz, he reports on world news, culture, and lifestyle. You can reach him at eric.mastrota@thenationaldigest.com.









