The company reports that the Omnii-1-designed phage variants eliminated the patient's antibiotic-resistant bacterial isolate in vitro, meaning in laboratory glassware rather than in a body. Beyond that, the trail stops: no results from the patient have been published, citing privacy, and the infection involved has not been named.

Radical Numerics is careful not to overclaim, noting that observations from a single patient establish little about whether the therapy is generally safe or effective. Phage therapy itself is not new; the bacteria-targeting viruses are already being explored as an alternative to antibiotics as drug-resistant infections spread. What is new is the AI design step reaching clinical administration.

There is precedent in the lab: in 2025, researchers at Stanford University and the Arc Institute created AI-designed phages targeting E. coli, several of which outperformed their natural counterpart in laboratory tests. Radical Numerics tested its model's ability to predict which phages could infect Pseudomonas aeruginosa, a bacterium that can resist antibiotics and causes wound and cystic fibrosis-related lung infections, and says it performed competitively against a specialised model. The write-up does not suggest the patient had a Pseudomonas infection.