● Covered by 4 sources · 4 reportsMedium impact1 negative3 neutral
AI Designs Novel Viruses Capable of Infecting Bacteria, Raising Biosafety Concerns
Key points
- AI models designed novel bacteriophages.
- Viruses killed antibiotic-resistant E. coli in tests.
- Models trained on genetic sequences, similar to LLMs.
- 16 out of 285 AI-generated genomes formed functional viruses.
- Research raises biosafety and biosecurity considerations.
AI-Designed Viruses Show Efficacy Against Bacteria
Scientists at Stanford University and the Arc Institute have successfully created the first viruses designed by artificial intelligence. These viruses, specifically bacteriophages, were developed using genome language models like Evo 1 and Evo 2, which function similarly to large language models but are trained on genetic sequences.
In laboratory tests, a cocktail of these AI-designed viruses demonstrated the ability to kill E. coli bacteria that were resistant to natural bacteriophages. This development suggests a potential for new medical treatments, particularly in addressing persistent infections caused by antibiotic-resistant bacteria.
How the AI Models Were Trained
The AI models were trained on trillions of nucleotides, the building blocks of DNA, to learn the statistical patterns and "grammar" of genetic code. For this specific experiment, the models were further trained on approximately 15,000 viruses from the same family as Phi X-174, a virus known to infect only E. coli.
The researchers explicitly limited the scope of the AI's design to exclude viruses that could infect humans, other animals, plants, or fungi. Out of 285 AI-generated viral genomes tested, 16 successfully assembled into functioning viruses capable of infecting and reproducing within bacteria.
Implications for Phage Therapy and Biotechnology
The ability to rapidly design and tune genomes for specific bacteria, while overcoming resistance, could transform phage therapy and expand biotechnological toolkits. Phage therapy, which uses viruses to treat bacterial infections, could benefit from AI's capacity to create targeted and effective viral agents.
This work represents an advancement in AI's capability to create biological entities, following previous research where large genome models generated DNA sequences encoding functional proteins and mimicking gene structures.
Biosafety and Biosecurity Concerns
While the potential benefits are significant, the scientists involved in the research have highlighted "important biosafety, biocontainment and biosecurity considerations." They urged others designing whole genomes to consult with relevant experts and consider the implications of such technology.
The development raises discussions regarding the governance of generative AI in genome design and the potential for misuse. Researchers suggest that preparations for the possibility of AI designing viruses targeting vertebrates should begin now.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
How outlets covered it
Researchers trained a genomic AI model, Evo, on trillions of nucleotides, enabling it to design complete DNA sequences for viruses. Of 285 AI-generated viral genomes tested, 16 successfully assembled into functioning viruses capable of infecting bacteria and reproducing. This development demonstrates AI's ability to generate novel biological entities, raising implications for synthetic biology and biosecurity.
Researchers at Arc Institute and Stanford University used AI models Evo 1 and Evo 2 to design novel viruses that can infect and reproduce within bacteria. This development demonstrates AI's capability to generate biological entities, raising both potential benefits for gene therapy and concerns about misuse.
Researchers at Stanford University have used large genome models to design the genomes of new viruses that infect bacteria. This development follows previous work where these models generated DNA sequences encoding functional proteins and mimicking gene structures, indicating an advancement in AI's capability to create biological entities.
Scientists have successfully created the first viruses designed by artificial intelligence, specifically bacteriophages, which demonstrated the ability to kill antibiotic-resistant E. coli in lab tests. This development offers potential for new medical treatments but also prompts significant biosafety and biosecurity discussions regarding the governance of generative AI in genome design.