AI Creates Novel Viruses to Target Bacteria Safely

WHO:

SFGATE - The study was published in the journal Science.

SFGATE - Researchers at Stanford University and the Palo Alto-based Arc Institute conducted the study.

SFGATE - Tom Ellis, a synthetic genome engineering professor at Imperial College London, commented on the research.

BBC - Brian Hie, assistant professor at Stanford University, commented on the use of generative AI in designing viruses.

WHAT:

SFGATE - The study revealed that researchers created nearly 300 unique viruses, of which 16 were deemed viable after successfully infecting bacteria.

SFGATE - The researchers instituted safeguards to prevent the AI model from creating harmful viruses by not providing information about viruses that infect humans or pose threats to plants and animals.

SFGATE - The study revealed that researchers created nearly 300 unique viruses, of which 16 were deemed viable after successfully infecting bacteria.

Wired - The use of AI systems to create viruses opens up new possibilities for combating bacterial resistance and raises concerns about the pace at which technology is outstripping regulation.

NY Post - Developing new bacteriophages could lead to new ways of treating infections that have become resistant to antibiotics, according to researchers.

Ars Technica - The AI system creates genetically distant versions of a bacteria-killing virus.

BBC - Brian Hie stated, "This is a next step in the complexity that's designable by generative AI, this is the first time generative AI has been used to design a complete genome, it's something that can replicate and have other functions inside cells… this was new territory for us."

Reddit Conspiracy - Scientists have created new viruses that multiply faster than natural ones.

Reddit Conspiracy - The AI system, named Evo, scanned about nine trillion nucleotides to generate codes for new genes that could perform specific tasks.

Techmeme - Scientists trained AI on genetic sequences to design 16 viable viruses that can infect bacteria but do not threaten humans.

BBC News - Scientists successfully created 16 viruses with genetic codes designed by artificial intelligence.

WHY:

SFGATE - The accompanying paper warned that the emerging technology has the potential to be abused, raising urgent biosafety and biosecurity questions.

HOW:

SFGATE - The AI model was trained on millions of natural genomes to produce specific outcomes, generating viruses modeled after a naturally occurring virus known as Phi X 174.

Techmeme - The AI model was trained on libraries of DNA and tasked with creating recipes for viral genomes.

BBC - The use of generative AI to design a complete genome represents a significant advancement in complexity.

(UPDATED August 7, 2026 22:28 UTC)

Researchers at Stanford University and the Palo Alto-based Arc Institute have published a study in the journal Science, revealing the creation of nearly 300 unique viruses, with 16 of them successfully infecting bacteria. To ensure safety, the researchers implemented safeguards in their AI model, preventing it from generating harmful viruses by excluding information about those that could infect humans or threaten plants and animals. However, the accompanying paper cautions that this emerging technology poses significant risks, raising urgent questions about biosafety and biosecurity.

(UPDATED August 7, 2026 19:48 UTC)

Researchers at Stanford University and the Arc Institute have developed an AI model that successfully created nearly 300 unique viruses, with 16 of them proving viable after infecting bacteria. To ensure safety, the researchers implemented safeguards to prevent the AI from generating harmful viruses by excluding information about those that could infect humans or threaten plants and animals. However, the accompanying paper cautioned that this emerging technology could be misused, raising important biosafety and biosecurity concerns.

(UPDATED August 7, 2026 15:48 UTC)

The use of AI systems to create viruses is opening up new possibilities for combating bacterial resistance. However, this development also raises concerns about the pace at which technology is outstripping regulation. Experts are calling for a closer examination of these advancements to ensure safety and ethical standards are maintained.

(UPDATED August 7, 2026 12:38 UTC)

Researchers are developing new bacteriophages that could provide innovative treatments for infections resistant to antibiotics. This breakthrough could significantly change the way we approach antibiotic-resistant infections. The findings were reported by the New York Post earlier today.

(UPDATED August 7, 2026 08:51 UTC)

A groundbreaking development in artificial intelligence has emerged, as an AI system has successfully created genetically distant versions of a bacteria-killing virus. Brian Hie noted that this marks a significant advancement in the complexity achievable through generative AI, highlighting that it is the first instance where such technology has been used to design a complete genome capable of replication and other cellular functions. This innovation represents a new frontier in the capabilities of AI in genetic design.

(UPDATED August 7, 2026 05:34 UTC)

Scientists have developed new viruses that multiply at a faster rate than those found in nature, according to a Reddit conspiracy discussion. An AI system named Evo played a crucial role in this process, scanning approximately nine trillion nucleotides to generate codes for new genes capable of performing specific tasks. This breakthrough raises significant questions about the implications of such advancements in genetic engineering.

(UPDATED August 7, 2026 00:49 UTC)

Scientists have successfully created 16 viable viruses using artificial intelligence, which were designed to infect bacteria without posing a threat to humans. The AI model was trained on extensive libraries of DNA and tasked with generating recipes for these viral genomes, marking a significant advancement in complexity. Brian Hie, an assistant professor at Stanford University, commented on this innovative use of generative AI in virus design.