Researchers in a Stanford laboratory have used artificial intelligence to design 16 new viruses that successfully kill bacteria, according to reports from RTL.nl and other outlets. The work marks the first time AI has produced viruses that do not occur naturally, as detailed in coverage from NRC, TPO.NL, HLN, and VRT.
The AI model generated the viruses, which are known as bacteriophages, and the designs proved functional when tested in the lab. Scientists involved in the project have described the achievement as highly significant, but they have also pointed to a dangerous side, according to VRT. The development has prompted calls for regulation, with NRC reporting that researchers are urging oversight of such AI capabilities.
The viruses were created in a Stanford lab, and the fact that they do not exist in nature sets them apart from previous efforts. While the immediate purpose is to target and destroy bacteria, the same technology could theoretically be applied to other types of viruses, raising concerns about misuse. The reports do not specify which bacteria were targeted or how the AI was trained, but the outcome is clear: the model produced 16 viable designs.
The news comes at a time when AI tools are increasingly being used in biology, and this case shows both the promise and the peril. On one hand, AI-designed phages could offer new ways to fight bacterial infections, especially those resistant to antibiotics. On the other hand, the ability to create novel viruses raises questions about biosafety and dual-use research.
What to watch next: whether regulators move to set rules for AI-driven virus design, and whether the Stanford team publishes details of its methods for broader scientific review.
