Scientists Use AI-Designed Viruses to Treat Human Patient
Scientists have administered artificial intelligence-designed viruses to a human patient for the first time in an experimental treatment aimed at fighting a serious bacterial infection, according to a report by The New York Times.
The treatment uses specially designed viruses called bacteriophages, or phages, which infect and kill bacteria. Researchers hope the AI-generated viruses can eliminate an E. coli infection that has resisted antibiotics and existing phage treatments.
The viruses were developed by US biotechnology startup Radical Numerics, which announced the treatment on Tuesday, October 20.
However, the company has not disclosed details about the patient or the doctors involved, citing confidentiality requirements.
How AI Designed the Viruses
Radical Numerics developed the treatment using its AI model, Omnii, which can analyze DNA sequences, protein structures, and other biological information.
The company created modified versions of a virus called T7, which specifically targets E. coli bacteria.
Researchers used Omnii to redesign the virus’s leg-like structures, allowing it to attach to bacterial strains that existing phages could not infect.
After generating the new viral genomes, scientists produced the corresponding DNA and inserted it into bacteria. Laboratory testing showed that the modified viruses could infect and kill the targeted bacteria.
The researchers then conducted safety checks before providing the viruses to doctors treating the patient.
FDA Authorizes Treatment
The US Food and Drug Administration (FDA) authorized the experimental treatment.
However, Radical Numerics has not revealed whether the patient has recovered or whether the treatment successfully cleared the infection.
The company’s technical findings have also not undergone peer review or been published in a scientific journal.
Earlier AI Research
The development builds on research conducted at the Arc Institute in California.
In 2025, researchers developed an AI model called Evo 2, trained on millions of DNA sequences, including those from viruses.
The model generated approximately 700,000 viral genomes. Researchers selected 285 designs for laboratory testing and found that 16 produced viable viruses.
Three researchers from the project, including Dr. Eric Nguyen, later established Radical Numerics.
In June 2026, the company introduced Omnii, an AI system designed to work with multiple forms of biological data.
Besides designing viruses, the company is exploring applications in cancer vaccine development and identifying potential pandemic risks.
Faster Treatment for Resistant Infections
Radical Numerics believes AI could significantly reduce the time required to develop personalized treatments for antibiotic-resistant infections.
Traditional phage therapy can involve lengthy searches for viruses capable of attacking a specific bacterial strain.
By using AI to design suitable viruses, researchers hope to shorten the process to a matter of days.
Eric Nguyen, the company’s chief executive, said the team believes it could eventually create personalized treatments within days.
However, researchers still need clinical evidence to establish whether AI-designed phages can safely and effectively treat infections in humans.
The treatment represents an early clinical application of AI-designed viruses, but its effectiveness remains unconfirmed.
The post Scientists Use AI-Designed Viruses to Treat Human Patient appeared first on ProPakistani.



