An AI-generated virus has been injected into a human. The announcement came on Tuesday. California-based AI laboratory Radical Numerics announced that an AI-designed virus has been administered to a human patient for the first time. The treatment, called Omnii-1, would, if verified, be the first instance of AI-designed bacteriophages (phages) being used in a patient, but the findings are expected to undergo scientific peer review.
Phages are viruses that infect bacteria. One estimate suggests that antibiotic-resistant superbugs will cause the deaths of approximately 40 million people by 2050. Current phage therapies proceed differently. In these treatments, researchers search large collections for a phage that matches the patient’s infection. Radical Numerics argues that it could speed up this process by rapidly producing phages on demand. Phages target only specific bacterial strains. This feature preserves beneficial gut microbes and constitutes the treatment’s critical advantage.
The company did not disclose, citing privacy. The company states that the model was trained on genome sequences and phage-host interaction data. The treatment was administered under a doctor’s supervision to a patient with an antibiotic-resistant infection whose medical options were running out. The company did not disclose clinical outcomes, citing patient privacy.
Why it matters
While antibiotic resistance is estimated to potentially cause tens of millions of deaths by 2050, phages have long been considered a source of hope for patients running out of treatment options; if this development is confirmed, it means phages can now be designed from scratch as needed, rather than having to be found in collections. Until now, finding the phage that matched each patient’s infection required a lengthy search process, and the claim of on-demand production targets precisely this bottleneck. However, it remains unknown whether the treatment administered to a single patient worked, because the company did not disclose clinical outcomes, citing privacy, and the findings have not yet undergone peer review. As a result, the announcement is less a milestone than a claim awaiting independent verification. The real question awaiting an answer is whether a virus designed by artificial intelligence is safe and effective in the human body.