A study has found artificial intelligence can generate virus designs that do not occur in nature. While that may sound concerning, scientists who advocate for the technology said the new viruses can unlock medical breakthroughs. 

The viruses in question were bacteriophages, meaning they target bacteria rather than people. Even so, outside experts have said the result is a significant sign of how much AI-driven biotechnology is advancing.

Here’s what to know

According to Smithsonian Magazine, citing research published in Science, the team used the generative AI models Evo 1 and Evo 2 to make new bacteriophages.

After being trained on vast amounts of genetic information, the models were asked to produce phages similar to Phi X-174, a widely studied virus that infects E. coli.

Out of roughly 700,000 candidate designs, the researchers selected 285 to build in the lab. Sixteen of those synthetic phages were able to halt E. coli growth in Petri dishes.

Those 16 AI-made phages also showed promise in later experiments: together, they successfully attacked two antibiotic-resistant E. coli strains. Phi X-174 and a mix of natural phages did not succeed in those same tests.

According to the World Health Organization, bacterial resistance was associated with more than 4.7 million deaths globally in 2021.

In comments to Carl Zimmer at The New York Times, as reported by The Smithsonian Magazine, Patrick Cai, a synthetic biologist at the University of Manchester who was not involved in the study, called it “an important milestone.”

More background

Researchers have long been interested in bacteriophages as a possible way to treat infections that no longer respond to antibiotics. If AI can speed up the search for effective phages, it could eventually help physicians target harmful bacteria more precisely.

That potential benefit comes with a risk. Technology that can create helpful viruses could also be abused if it were directed toward dangerous pathogens.

Isaac Bogoch, an infectious diseases specialist at the University of Toronto who was not involved in the research, told John Power and Erin Hale at Al Jazeera, as quoted by Smithsonian Magazine, that “A.I.-designed viruses could have some potential benefits, such as the creation of targeted bacteriophages that could possibly help us tackle antibiotic-resistant infections in new ways.”

There are technical limitations as well. Tom Ellis, a synthetic genome engineer at Imperial College London who was not involved in the study, told Al Jazeera that the phage used in this research has an extremely small genome, meaning more complex viruses are much harder to design.

What’s being done?

According to Smithsonian Magazine, the researchers said they tried to limit the danger by leaving certain genomes out of Evo’s training data, preventing the system from generating viruses that could infect humans, animals, plants, or fungi.

They also said they adopted safety measures beyond what is standard in bacteriophage research and outlined those steps in the paper’s additional materials as a possible biosafety framework for future work.

Another approach could involve using mixed phage therapies instead of relying on a single virus. Study co-author Brian Hie, a computational biologist at Stanford University, said in a statement that pairing genetically distinct phages could make it harder for bacteria to evolve around treatment.

As the technology develops, oversight, screening, and clear limits may need to keep pace.

Get TCD’s free newsletters for easy tips, smart advice, and a chance to earn $5,000 toward home upgrades. To see more stories like this one, change your Google preferences here.