---
title: "AI-designed bacteriophages created at Arc Institute infect E. coli"
publication: "MediaBias News"
url: "https://mediabias.news/tech/ai-designed-bacteriophages-created-at-arc-institute-infect-e-coli"
api: "https://mediabias.news/api/v1/stories/ai-designed-bacteriophages-created-at-arc-institute-infect-e-coli"
markdown: "https://mediabias.news/tech/ai-designed-bacteriophages-created-at-arc-institute-infect-e-coli.md"
audio: "https://mediabias.news/api/audio/ai-designed-bacteriophages-created-at-arc-institute-infect-e-coli"
category: "Science & Tech"
area: "United States"
published: "2026-08-06T22:25:30.328Z"
source_reported: "2026-08-06T22:02:40.429Z"
updated: "2026-08-07T00:45:10.900Z"
trust_score: 96
critic_score: 95
hype_score: 45
assessment_type: "coverage-and-source-reporting-analysis"
fact_check_status: "not-performed"
coverage_measured: "2026-08-07T00:45:06.964998+00:00"
---

# AI-designed bacteriophages created at Arc Institute infect E. coli

*Researchers generated 700,000 candidate genomes before 16 proved viable in lab tests*

**Scores for the source reporting** (0-100, assessing the original journalism this article was written from, not this write-up): 

- Trust 96 of 100, higher is better. Trust high (96) due to named researchers, published Science study, and multiple independent outlets reporting details.
- Craft 95 of 100, higher is better. Critic high (95) as articles explain significance, quote experts, give specific figures, and outline next steps.
- Hype 45 of 100, lower is better. Hype moderate (45) with many sensational headlines but bodies remain factual and measured.

Scored by MediaBias News; method at https://mediabias.news/methodology.

**The short version**

- Researchers at the Arc Institute used an AI model to design 16 viable bacteriophages that infect E. coli, as reported in Science on 6 August 2026.
- The AI generated 700,000 candidate genomes before narrowing to 300 tested in the lab, of which 16 proved functional.
- The study excluded human pathogen datasets, ensuring the viruses cannot infect people, but biosecurity experts warn of potential misuse.
- The National Institutes of Health’s new policy does not restrict AI-based research unless it involves an existing entity of concern, such as smallpox.

Researchers at the Arc Institute in Palo Alto, California, used an artificial intelligence model named Evo to design 16 new bacteriophages capable of infecting E. coli. The study, published in the journal Science on 6 August 2026, trained the AI on DNA sequences from naturally occurring viruses before tasking it to generate novel viral genomes. The team synthesised 300 of the 700,000 candidate genomes produced by Evo, and 16 proved viable in laboratory tests.

The viruses are bacteriophages, which infect bacteria but cannot infect humans. The study excluded datasets containing human pathogens from the AI’s training, ensuring the designed viruses pose no direct threat to people. The 16 viable phages overcame resistance in two strains of E. coli, demonstrating functionality beyond simple imitation of natural viruses. The research was conducted by a team including scientists from Stanford University.

An accompanying article in Science, authored by biosecurity experts at Johns Hopkins University, warned that the ability to design viral genomes using AI raises urgent biosafety and biosecurity questions. The National Institutes of Health issued new policies last week to prevent high-risk life science research but did not restrict AI-based research unless it involves an existing entity of concern, such as smallpox.

## What the coverage left out

None of the right-rated digests mentioned the potential medical applications of the research, such as combating antibiotic-resistant bacteria. Neither of the right-rated digests carried the biosecurity warnings from Johns Hopkins experts or the fact that the AI generated 700,000 candidate genomes before producing 16 viable viruses. The left-rated digests did not mention the potential for medical breakthroughs.

## Who covered it

Shares of the 24 covering outlets with a published leaning rating:

- Left: 21% (8)
- Centre: 58% (12)
- Right: 21% (4)

11 of 35 covering outlets carried no usable leaning rating and were excluded from those percentages. Coverage measured 2026-08-07T00:45:06.964998+00:00.

**Coverage watch:** developing. Checked 12 times; 1 check found a material change. Most recently checked 2026-08-07T01:15:06.535Z.

## How the sides framed it

### Left

- The left-rated outlets led on the first successful creation of synthetic viruses designed entirely by artificial intelligence. All six digests reported that 16 AI-generated viral genomes proved viable, capable of infecting and replicating in bacteria. The research was attributed to a Stanford-led team at the Arc Institute.
- Five of the six digests described the process: AI was trained on DNA libraries and then prompted to generate novel viral genomes. None of the digests omitted the number of viable viruses produced. The Sydney Morning Herald, The Age, and Brisbane Times carried identical accounts of the method and outcome.
- Two digests highlighted potential risks. Axios framed the development as a step toward a scenario where AI-designed pathogens outpace surveillance. unn.ua noted the technology’s potential for medicine and biotechnology, while IFLScience quoted a researcher emphasising the positive impact of advancing such models.

### Centre

- The centre-rated reports balanced the potential benefits and risks of the research. Forbes’s full report led on the AI’s ability to invent 16 new viruses and quoted a biosecurity expert warning that the technology could be used to create bioweapons. The outlet also highlighted the potential for breakthroughs in treating antibiotic-resistant superbugs, noting that the viruses could be used to generate tailor-made therapies.
- The Times’s digest framed the story as raising "hopes" for tackling drug-resistant infections but also "urgent biosafety questions". The centre-rated digests uniformly carried the fact that the viruses were designed to infect bacteria and were harmless to humans. Most mentioned the 700,000 candidate genomes and the 16 viable viruses, but none led on the number of attempts required. BBC News’s digest and ANSA’s digest both noted the dual potential for medical advances and misuse.

### Right

- The right-rated digests focused on the achievement itself and the safety assurances. Thairath’s digest stated that the viruses were "designed to kill bacteria and are harmless to humans", without mentioning biosecurity concerns. ABC’s digest described the research as a "new leap" in creating "complete artificial organisms" and framed it as the beginning of a "brilliant, although also uncertain" future, but did not quote biosecurity warnings or the potential for medical applications.
- Neither of the right-rated digests mentioned the 700,000 candidate genomes or the 16 viable viruses. Both carried the fact that the viruses cannot infect humans, but neither included the biosecurity warnings from Johns Hopkins experts or the potential for the technology to be misused.

## Factuality profile of the covering outlets

These are published factuality ratings of the outlets, not a verdict on whether this story or its claims are true.

- mixed: 3
- high: 17
- unknown: 10
- veryHigh: 5

## Verification scope

This page compares coverage and scores the source reporting. It is not a ClaimReview verdict on whether the underlying event or claim is true.

## Original reporting this was written from

- [KIFI](https://localnews8.com/health/cnn-health/2026/08/06/ai-creates-16-new-viruses-from-scratch-showing-promise-for-drug-resistance-and-drawing-warnings-about-potential-for-misuse) — AI creates 16 new viruses from scratch, showing promise for drug resistance and drawing warnings about potential for misuse
- [KRDO](https://krdo.com/news/2026/08/06/ai-creates-16-new-viruses-from-scratch-showing-promise-for-drug-resistance-and-drawing-warnings-about-potential-for-misuse) — AI creates 16 new viruses from scratch, showing promise for drug resistance and drawing warnings about potential for misuse
- [KEYT](https://keyt.com/health/cnn-health/2026/08/06/ai-creates-16-new-viruses-from-scratch-showing-promise-for-drug-resistance-and-drawing-warnings-about-potential-for-misuse) — AI creates 16 new viruses from scratch, showing promise for drug resistance and drawing warnings about potential for misuse
- [Sydney Morning Herald](https://smh.com.au/world/north-america/ai-used-to-create-new-viruses-in-the-lab-20260807-p60m7g.html) — AI used to create new viruses in the lab
- [The Age](https://theage.com.au/world/north-america/ai-used-to-create-new-viruses-in-the-lab-20260807-p60m7g.html) — AI used to create new viruses in the lab
- [Brisbane Times](https://brisbanetimes.com.au/world/north-america/ai-used-to-create-new-viruses-in-the-lab-20260807-p60m7g.html) — AI used to create new viruses in the lab
- [WA today](https://watoday.com.au/world/north-america/ai-used-to-create-new-viruses-in-the-lab-20260807-p60m7g.html) — AI used to create new viruses in the lab
- [PressNewsAgency](https://pressnewsagency.org/ai-used-to-create-new-viruses-in-the-lab) — AI used to create new viruses in the lab
- [The Independent](https://independent.co.uk/news/science/ai-virus-dna-stanford-ecoli-b3028990.html) — Nothing to fear: AI just created a virus that is not found in nature
- [Western Morning News](https://westernmorning.news/2026/08/06/nothing-to-fear-ai-just-created-a-virus-that-is-not-found-in-nature) — Nothing to fear: AI just created a virus that is not found in nature
- [Crypto Briefing](https://cryptobriefing.com/ai-designs-functional-viral-genomes-bacteriophages) — AI designs functional viral genomes from scratch, and 16 of them actually work
- [Unknown](https://chosun.com/economy/science/2026/08/07/GBQTCODBMM2DEMDBG43TONRTGI) — AI Creates Viruses That Kill Bacteria, Infecting Even Resistant Bacteria
- [unn.ua](https://unn.ua/news/shtuchnyi-intelekt-upershe-stvoryv-novi-zhyttiezdatni-virusy-yakykh-nemaie-v-pryrodi) — Artificial Intelligence Has for the First Time Created New Viable Viruses that Do Not Exist in Nature
- [The Times](https://thetimes.com/uk/science/article/ai-designed-viruses-biosafety-2s9jgn66s) — World’s first AI-designed viruses raise hopes
- [Internewscast](https://internewscast.com/news/ai-designed-viruses-are-here-the-promise-and-peril-ahead) — AI-Designed Viruses Are Here: The Promise and Peril Ahead
- [thairath.co.th](https://thairath.co.th/news/foreign/2951136) — Scientists Have Successfully Used AI to Design and Create a New Type of Virus for the First Time.
- [Axios](https://axios.com/2026/08/06/ai-virus-designed-bacteria-viruses) — AI designs new virus not found in nature
- [Gizmodo](https://gizmodo.com/oh-cool-ai-is-inventing-new-biological-viruses-now-2000795623) — Oh Cool, AI Is Inventing New Biological Viruses Now
- [BIOENGINEER.ORG](https://bioengineer.org/ai-designs-functional-bacteriophages-entirely-from-scratch) — AI Designs Functional Bacteriophages Entirely From Scratch
- [Giornale di Sicilia](https://gds.it/articoli/cronaca/2026/08/06/intelligenza-artificiale-puo-creare-nuovi-virus-5d678bbc-f871-4ae0-9535-761c8ddf513f) — • Artificial Intelligence Can Create New Viruses • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • Artificial Intelligence Can Create New Viruses
- [Il Messaggero](https://ilmessaggero.it/salute/focus/virus_intelligenza_artificiale_scienziati_pericoli_cosa_succede-9694331.html) — New Viruses Created by Scientists Thanks to Artificial Intelligence. It Is the First Time in History.
- [Ars Technica](https://arstechnica.com/science/2026/08/large-genome-models-used-to-design-new-viruses) — Large genome models used to design new viruses
- [Adevarul](https://adevarul.ro/stil-de-viata/stiinta/temeri-privind-siguranta-oamenii-de-stiinta-au-2548238.html) — Major Health Safety Concerns: Scientists Have Created Artificial Intelligence Viruses
- [A TARDE](https://atarde.com.br/tecnologia/inteligencia-artificial-projeta-virus-ineditos-em-laboratorio-pela-primeira-vez-1397769) — Artificial Intelligence Projects Unpublished Viruses in the Lab for the First Time
- [la Nacion](https://lanacion.com.ar/sociedad/esta-ia-acaba-de-crear-virus-que-no-se-encuentran-en-la-naturaleza-nid06082026) — This AI Just Created Viruses that Are Not Found in Nature
- [ANSA](https://ansa.it/sito/notizie/topnews/2026/08/06/nyt-scienziati-usano-lia-per-creare-nuovi-virus-e-la-prima-volta_6d4dfb7e-fd1d-4b6c-a2a0-2c886c5d153e.html) — Nyt, 'Scientists Use AI to Create New Viruses, It's the First Time'
- [yle](https://yle.fi/a/74-20239791) — Artificial Intelligence Just Created Viruses that Are Not Found in Nature
- [Heise](https://heise.de/news/KI-Modelle-entwerfen-lebensfaehige-Viren-Forscher-warnen-vor-Missbrauch-11402252.html) — AI Models Design Viable Viruses, Researchers Warn Against Abuse
- [abc](https://abc.es/ciencia/ia-gran-salto-disenar-genes-crear-organismos-20260806203647-nt.html) — AI Takes a New Leap and Creates Complete Artificial Organisms From Scratch
- [Science Media Centre](https://sciencemediacentre.org/expert-reaction-to-generative-design-of-bacteriophages-with-genome-language-models) — expert reaction to generative design of bacteriophages with genome language models
- [BBC News](https://bbc.com/news/articles/c5y3j3ngevmo) — Artificial Intelligence used to design brand new viruses
- [Unknown](https://kumanichi.com/articles/2022884) — Generating AI Designs "New Virus"; Confirmed to Grow in Bacteria, US University Research (Kyodo News)
- [IFLScience](https://iflscience.com/in-a-world-first-scientists-used-ai-to-generate-a-genome-and-produced-a-functional-bacteria-killing-virus-84304) — In A World First, Scientists Used AI To Generate A Genome And Produced A Functional, Bacteria-Killing Virus
- [oem.com.mx](https://oem.com.mx/elsoldemexico/ciencia-y-salud/la-ia-ya-es-capaz-de-crear-virus-que-no-existen-en-la-naturaleza-a-partir-del-adn-de-otros-patogenos-describe-science-31461658) — AI Is Already Able to Create Viruses that Do Not Exist in Nature From the DNA of Other Pathogens, Describes Science - El Sol De México
- [Forbes](https://forbes.com/sites/maryroeloffs/2026/08/06/scientists-trained-an-ai-model-in-dna-and-it-invented-16-new-viruses) — Scientists Trained An AI Model In DNA—And It Invented 16 New Viruses

## Questions

**How many viruses did the AI design that proved viable?**

The AI model Evo generated 700,000 candidate genomes, which researchers narrowed to 300 tested in the lab. Of those, 16 proved viable and capable of infecting E. coli, as reported in the study published in Science on 6 August 2026.

**Can the AI-designed viruses infect humans?**

No. The study excluded human pathogen datasets from the AI’s training, ensuring the designed viruses are bacteriophages that can only infect bacteria. The researchers confirmed the viruses are not capable of infecting people, as reported by Gizmodo and Forbes.

**What are the biosecurity concerns raised by this research?**

Biosecurity experts at Johns Hopkins University warned in Science that the ability to design viral genomes using AI raises urgent biosafety and biosecurity questions. They called for strict laws to govern future research, particularly to prevent the use of similar techniques on pathogens affecting humans, animals or crops.

**Does the National Institutes of Health’s new policy restrict AI-based virus research?**

No. The National Institutes of Health’s new policy bars scientists from conducting experiments that would make biological agents more dangerous but does not restrict AI-based research unless it involves an existing entity of concern, such as smallpox, as reported by Gizmodo.

---

MediaBias News — https://mediabias.news. Reproduced from https://mediabias.news/tech/ai-designed-bacteriophages-created-at-arc-institute-infect-e-coli. Please cite as: MediaBias News, "AI-designed bacteriophages created at Arc Institute infect E. coli", https://mediabias.news/tech/ai-designed-bacteriophages-created-at-arc-institute-infect-e-coli