This specific study sounds relatively benign, but generative models for DNA sequences sounds like an incredibly dangerous technology to develop. If someone were to reproduce their study while including animal viruses in the training data, what would it be capable of? And what about after a few years of progress?
Their phages have only 5400 base pairs, but the hepatitis virus has around 3000, and ebola only had ~19,000. That doesn't mean making a novel lethal virus is easy, but if anyone pulls it off, it would make Covid look like a picnic.
I have been diving into / reading up this area lately. A few quick notes.
Worth specifying that these were hand selected candidates. Out of these human selected ones, 5% were successful.
The Stanford researchers picked the most promising 302 AI designs and synthesised them in the lab. Of these, 16 proved effective at killing E. coli bacteria.
"picked the most promising 302 AI designs" out of a n that isn't mentioned in this article.
Based on my quick reading, and I hope I'm wrong here / missed something, but it feels like the BBC didn't use a search engine.
Because this isn't a "first." Not by a long shot. And it's not even a first by the same team!
From what I can tell, each reports iteration on the project / progress. And it's important to say that their work is excellent and I really like their paper (I'm reading it right now)...
But it feels like they've iterated through titles / angles for the same work until they found the viral version.
More generally, re: algorithm builds life. There's an entire sub-field / world here that goes back decades at this point. The first synthetic organism was Venter's mycoplasma laboratorium / Synthia in 2010.
The first organism to be synthetically assembled (not designed, but assembled ikea style) was a phage in 2003 also by Venter & co.
In 2021 another team made entities that can "find and combine building blocks into self-copies," Kinematic self-replication in reconfigurable organisms
https://pmc.ncbi.nlm.nih.gov/articles/PMC8670470/
> Based on my quick reading, the BBC article doesn't bring up a prior publication / release from 2025 called "AI-Designed Phages," https://www.asimov.press/p/ai-phages
That link is a commentary about the preprint of the article mentioned in the BBC article.
All those publications describe the methodology and modelling process to generate genomes, without testing them experimentally. This article shows that their designed genomes actually work in the wetlab. Your last two links are both preprints.
That's a fair point! I think I've found a few more things, but reading through / verifying first. Will update accordingly and leave notes. Thanks for helping!
This specific study sounds relatively benign, but generative models for DNA sequences sounds like an incredibly dangerous technology to develop. If someone were to reproduce their study while including animal viruses in the training data, what would it be capable of? And what about after a few years of progress?
Their phages have only 5400 base pairs, but the hepatitis virus has around 3000, and ebola only had ~19,000. That doesn't mean making a novel lethal virus is easy, but if anyone pulls it off, it would make Covid look like a picnic.
I have been diving into / reading up this area lately. A few quick notes.
Worth specifying that these were hand selected candidates. Out of these human selected ones, 5% were successful.
"picked the most promising 302 AI designs" out of a n that isn't mentioned in this article.Based on my quick reading, and I hope I'm wrong here / missed something, but it feels like the BBC didn't use a search engine.
Because this isn't a "first." Not by a long shot. And it's not even a first by the same team!
Same team, prior publication / release from 2025 called "AI-Designed Phages," https://www.asimov.press/p/ai-phages
edit - note this wasn't a pre-print, the same work has been (re)-published multiple times / been floated multiple times in the past 18 months,
"Genome modelling and design across all domains of life with Evo 2," March 2026, https://www.nature.com/articles/s41586-026-10176-5
"Generative design of novel bacteriophages with genome language models," September 2025, https://www.biorxiv.org/content/10.1101/2025.09.12.675911v1
Even older conference paper, I think, February 19th, 2025, https://arcinstitute.org/manuscripts/Evo2
From what I can tell, each reports iteration on the project / progress. And it's important to say that their work is excellent and I really like their paper (I'm reading it right now)...
But it feels like they've iterated through titles / angles for the same work until they found the viral version.
More generally, re: algorithm builds life. There's an entire sub-field / world here that goes back decades at this point. The first synthetic organism was Venter's mycoplasma laboratorium / Synthia in 2010.
The first organism to be synthetically assembled (not designed, but assembled ikea style) was a phage in 2003 also by Venter & co.
In 2021 another team made entities that can "find and combine building blocks into self-copies," Kinematic self-replication in reconfigurable organisms https://pmc.ncbi.nlm.nih.gov/articles/PMC8670470/
From 2019 / 2020, Google / Alphabet made algorithmically designed synthetic life / "xenobots", https://www.nytimes.com/2020/04/03/science/xenobots-robots-f... https://en.wikipedia.org/wiki/Xenobot
Zurich in 2019, https://web.archive.org/web/20191008040536/https://www.eurek...
And if you expand computing to include molecular processes, there's stuff like PACE, https://en.wikipedia.org/wiki/Phage-assisted_continuous_evol...
> Based on my quick reading, the BBC article doesn't bring up a prior publication / release from 2025 called "AI-Designed Phages," https://www.asimov.press/p/ai-phages
That link is a commentary about the preprint of the article mentioned in the BBC article.
After your editing:
> Same team, prior publication / release from 2025 called "AI-Designed Phages," https://www.asimov.press/p/ai-phages
It is the same work, just not published in Science.
Apologies was editing the comment, but I looked into it more and it wasn't a preprint. They published it. Multiple times.
"Genome modelling and design across all domains of life with Evo 2," March 2026, https://www.nature.com/articles/s41586-026-10176-5
However, this was a preprint, "Generative design of novel bacteriophages with genome language models," September 2025, https://www.biorxiv.org/content/10.1101/2025.09.12.675911v1
Even older conference paper, I think, February 19th, 2025, https://arcinstitute.org/manuscripts/Evo2
All those publications describe the methodology and modelling process to generate genomes, without testing them experimentally. This article shows that their designed genomes actually work in the wetlab. Your last two links are both preprints.
That's a fair point! I think I've found a few more things, but reading through / verifying first. Will update accordingly and leave notes. Thanks for helping!
The publication: https://www.science.org/doi/10.1126/science.aec2657
It is paywalled, but this seems to be the corresponding preprint: https://www.biorxiv.org/content/10.1101/2025.09.12.675911v1....
Anyone else assume this was about computer viruses and exploits...
Now it just sticks to using the old ones?