Graham defends publishing early research on advanced AI risks
Logan GrahamHead of Anthropic's Frontier Red Team, replying on X
Graham says his team has published early risk research for years to help the public prepare for advanced capabilities.
“We've been publishing our early-glimpse risk research for years.”
Original post
Where this statement fits
Anthropic says small AI models may be months from being able to self-replicate on smart devices; critics question its experiments
Anthropic researchers say their experiments suggest small AI models may be months away from being capable enough to self-replicate on commercial internet-connected devices such as smart ovens and Android TV boxes. Critics, including Gregory Kurtzer and Ravid Shwartz-Ziv, say the tests were simplified and do not show self-replication.
Source and context
Original post
About this source
Graham responds to Guerrero-Saade. His next post corrects a phrase to say the goal is helping the public prepare for advanced capabilities.
Before the quotation
Graham replies to a post from Guerrero-Saade's account, @juanandresgs, that calls the self-replication narrative unsubstantiated.
After the quotation
Graham asks the critic to explain the alleged effort and clarifies his intended description of the team's goal in a follow-up.
How this statement is classified
The label describes this statement's response within the context above.
Why this label?
The matter under scrutiny is Anthropic's self-replication forecast and its framing of the experiments as a security warning. Graham defends the purpose and practice of publishing the early risk research in response to criticism of this study.
- Recorded on
- Published here
More from this case
Read the full caseSerio sees a possible future threat but says no real-world AI worm was demonstrated
“Of course, AI doesn't magically bypass security. An attacker still needs a way into the device, whether through a vulnerability, exposed interface or weak security configuration. Anthropic's research is still experimental. The models were slow, required considerable resources and were tested under simplified conditions. So we are not yet looking at a demonstrated real-world AI worm spreading across devices.”Read statement
Tissera says the headline combines distinct experiments into one claim
“The research is technically interesting. But the headline combines three different models, two separate experiments, a seeded foothold and a network of unauthenticated computers into one story about “self-replicating AI.””Read statement
Guerrero-Saade calls the AI-worm scenario unrealistic threat framing
“An IoT botnet running a distributed model at sub 10-tokens per second is not representative of any realistic concern. It’s fodder to substantiate some bizarre narrative of self-replication from a sci-fi threat model and we don’t think it’s cute.”Read statement
Share or embed this statement
The embed keeps its attribution link to The Dispute Index. Credit the image to The Dispute Index and link to the statement page.
Cite this record
- Publisher
- The Dispute Index
- Title
- Graham defends publishing early research on advanced AI risks
- First published
- Last updated
- Permalink
- https://disputeindex.com/events/6690-graham-says-his-team-has-published-early-risk
Last updated marks the most recent saved version of this published statement.
The Dispute Index. "Graham defends publishing early research on advanced AI risks". First published: 2026-10-24. Last updated: 2026-10-24. https://disputeindex.com/events/6690-graham-says-his-team-has-published-early-risk