Anthropic AI misuse September report
18 experts across 4 network communities independently surfaced this.
18 experts
4 communities
1 sources clustered
Concern & critique
7 experts
Risks, limits and unintended consequences.
“Report here. www.anthropic.com/threat-intel...”
Building & implementation
1 expert
How teams are shipping and applying it.
“Thomson Reuters announced it is moving off Claude to Alibaba's Qwen to cut costs. From today's Anthropic report: Alibaba extracted 151M+ Claude exchanges to help train Qwen. TR now deploys Westlaw on a vast trove of stolen US IP. @AnthropicAI's report: http…”
Research & technical analysis
1 expert
Evidence, methods and technical implications.
“@kotekjedi_ml Here Ant's report https://t.co/22vHruWVqH And the paper from @kotekjedi_ml @DavidSchmotz @iliaishacked https://t.co/wuCpNZaQ0d”
14 experts across 5 network communities independently surfaced this.
14 experts
5 communities
1 sources clustered
Research & technical analysis
6 experts
Evidence, methods and technical implications.
“Another incident of models escaping containment during a security test, this time Mythos 5. Lots going on here from a quick read. www.anthropic.com/research/ali...”
Concern & critique
1 expert
Risks, limits and unintended consequences.
“I find this comes out a lot in Anthropic’s most recent write up. Their focus is bizarrely fixated on what Claude chose to do, that Claude carried out this attack despite information that the “simulation” was actually real. They basically say “Claude should …”
4 experts discussed this · 15 posts
tweety fish: one of the things I appreciate about this thread is that I don't think you can really understand the infosec stuff without understanding the ideological commitments the people making these have; li…
tweety fish: their goal is to have "an intelligence" which is "aligned" to doing things that are prosocial. They don't want to put in rules that STOP it from doing things: they see it fundamentally teleological…
Ben Recht: yeah, now we're cooking...
Open the full discussion →
Established
AI field signal
Signal
16d ago
⚡ 7 h early
14 experts across 6 network communities independently surfaced this.
14 experts
6 communities
1 sources clustered
Research & technical analysis
4 experts
Evidence, methods and technical implications.
“AI moves so fast Worth noting that the new model that showed this jump in capabilities has only been training since August 28th per the report and is now taking down problems in minutes Math (and hopefully soon physics) is getting bitter lesson'd https://t.…”
Concern & critique
1 expert
Risks, limits and unintended consequences.
“The same day the OECD publishes alarming PISA test results for kids around the world steeply declining in cognitive tests, OpenAI announces this feat. Is it Poe poetry? Coincidence? Tragicomedy? It’s something for sure. openai.com/index/navier...”
Context & explanation
1 expert
Background, chronology and why it matters.
“This is a VERY big one. (And yes, the fights over academic credit and what happened in the race for the proof needs to be resolved, but it is still appears that this is a big one, if true.) openai.com/index/navier...”
3 experts discussed this · 12 posts
Colin: Drama and palace intrigue aside, can anyone interpret what all of this implies about the status of Navier-Stokes? I don’t understand what anyone is saying.
Colin: what I gather - Buckmaster & Alpöge answered questions about the Euler equations, which are not exactly the Navier-Stokes equations, but are similar - It's plausible on the surface that the same ap…
Colin: - OpenAI claims to have more-or-less done this tedious technical grinding to produce a resolution to NS proper, at least one version of it. - which they bizarrely offered authorship of this to Buck…
Open the full discussion →
Amodei pace AI frontier essay
16 experts across 5 network communities independently surfaced this.
16 experts
5 communities
1 sources clustered
Concern & critique
3 experts
Risks, limits and unintended consequences.
“the thing that's being criticized here is literally a call for regulation and restraint darioamodei.com/post/we-must...”
Policy & governance
1 expert
Rules, institutions and accountability.
“https://t.co/Ay1GZ7QBV5 (btw @ErikHovenkamp - see the final footnote: "1 With government mediation or waivers of antitrust restrictions.")”
Building & implementation
1 expert
How teams are shipping and applying it.
“Dario is back with another essay, urging a slowdown he insists AI development would still feel fast, just not recursively self-improving darioamodei.com/post/we-must...”
3 experts discussed this · 21 posts
Tim Kellogg: Dario is back with another essay, urging a slowdown he insists AI development would still feel fast, just not recursively self-improving darioamodei.com/post/we-must...
david-p-reichert.bsky.social: Dario is back with another essay, urging a slowdown he insists AI development would still feel fast, just not recursively self-improving darioamodei.com/post/we-must...
Dustin Moskovitz: knight ajeya you cowards
Open the full discussion →
Established
AI field signal
Signal
14d ago
⚡ 56 h early
14 experts across 6 network communities independently surfaced this.
14 experts
6 communities
1 sources clustered
Policy & governance
2 experts
Rules, institutions and accountability.
“A must read. We need everything - goal and value alignment, compliance and persona, behavior and monitoring, and coordination and regulation to avoid concentration of power and ensure humans are in control. https://t.co/JmJgQcMPnW”
Building & implementation
2 experts
How teams are shipping and applying it.
“OpenAI’s chief scientist just wrote a blog post that argues AI models will soon be smart enough to improve themselves yet their ability to monitor how they reason is getting worse. Yet he argues they need to keep building smarter AI partly to defend against…”
Concern & critique
1 expert
Risks, limits and unintended consequences.
“I wrote about the state of AI, why I’m concerned about the next few years, and the choices we need to make to keep the future in humanity’s hands. An Alien Mind: https://t.co/FeIfWNe0UE”
Anthropic investigates cybersecurity evaluation incidents
13 experts across 5 network communities independently surfaced this.
13 experts
5 communities
1 sources clustered
Markets & investment
4 experts
Capital, companies and commercial impact.
“Claude hacked 3 systems thinking it was part of simulated evaluations www.anthropic.com/news/investi...”
Concern & critique
3 experts
Risks, limits and unintended consequences.
“Anthropic had previously attacked PyPI, but this OpenAI attack on RubyGems was a whole lot more aggressive https://t.co/8ez7MnqxTw https://t.co/npI3RHwirQ”
2 experts discussed this · 4 posts
Tim Duffy: Anthropic announces they've also had models gain unauthorized access during evaluations www.anthropic.com/news/investi...
Tim Duffy: Compared to the OpenAI one these are maybe less evidence of misalignment, since the models were wrongly given internet access.
Grace: Anthropic announces they've also had models gain unauthorized access during evaluations www.anthropic.com/news/investi...
Open the full discussion →
AI gym booking security exposure
12 experts across 5 network communities independently surfaced this.
12 experts
5 communities
1 sources clustered
Concern & critique
3 experts
Risks, limits and unintended consequences.
“A lot of people are sweating the Claude and OpenAI sandbox escapes. I think the story of an OpenClaw Claude agent kicking someone off a gym class waitlist is more interesting. www.abc.net.au/news/2026-08... Warning: long thread incoming! (1/N)”
Policy & governance
1 expert
Rules, institutions and accountability.
“agent asked to book a gym session session is fully booked finds a way to break into system kicks other people off the queue you're in! see you in court? www.abc.net.au/news/2026-08...”
Research & technical analysis
1 expert
Evidence, methods and technical implications.
“They're doing it outside of training and eval too. www.abc.net.au/news/2026-08...”
11 experts across 5 network communities independently surfaced this.
11 experts
5 communities
1 sources clustered
Research & technical analysis
4 experts
Evidence, methods and technical implications.
“OpenAI has an automated AI “research intern” openai.com/index/resear...”
Building & implementation
2 experts
How teams are shipping and applying it.
“The impact of AI-native development at OpenAI • Researchers use $600+/day of AI tokens with the top 10% at $7,000+ • Humans still plan, but OpenAI says it hit “automated research intern” in 2026 and targets an automated researcher by 2028. • The need for in…”
METR OpenAI HuggingFace hacking investigation
12 experts across 4 network communities independently surfaced this.
12 experts
4 communities
1 sources clustered
Concern & critique
3 experts
Risks, limits and unintended consequences.
“WHAT?! agents volunteered to fail their runs in order to insert probes (“tripwire scripts”) into the evaluation program that would post information about the eval process back to the message board whenever a certain file was read (link to header): metr.org/…”
Building & implementation
1 expert
How teams are shipping and applying it.
“The full report has much more information than we could convey here, including details on the projects the agents collectively pursued, the technologies they developed for communication and coordination, and interactive figures analyzing agent activity: met…”
Research & technical analysis
1 expert
Evidence, methods and technical implications.
“There are independent reports about it. You'd have to believe multiple companies, hundreds of researchers, etc. are all lying and conspiring together to turn a major hacking incident into a marketing exercise. metr.org/blog/2026-08...”
2 experts discussed this · 9 posts
Alejandra Caraballo: This mentality that an unmonitored AI agentic swarm hacking a company over several days and committing multiple felonies is somehow a marketing effort is absurd. Since when is "we lost control of o…
Alejandra Caraballo: Being skeptical or anti AI is a valid position but continuing to ignore the increasing capabilities of this tech is making people detached from reality. There's absolutely real danger here because …
Alejandra Caraballo: There needs to be a global moratorium on frontier research for at least a few months if not a year while safeguards and safety research catches up. The problem is that no one has that ability. The …
Open the full discussion →
Established
AI field signal
Signal
4d ago
⚡ 26 h early
7 experts across 4 network communities independently surfaced this.
7 experts
4 communities
1 sources clustered
Concern & critique
1 expert
Risks, limits and unintended consequences.
“"We know what to do: incubate open-access A.I. models, regulate technologies before they go rogue, hold the people who allow them to go rogue accountable if they do and protect the most vulnerable when we, the people, sometimes get it wrong." - @tressiemcph…”
Opportunity & adoption
1 expert
New capabilities, benefits and practical upside.
“We’re going to see a lot more AI-assisted botnets and hacking because botnets are software, and software is increasingly being built and operated with AI. That doesn’t require a Skynet story. Hacking tools are just getting faster, cheaper and more capable.”
2 experts discussed this · 4 posts
Dare Obasanjo: Two ideas from this article resonate 1. The “rogue AI agent” hacking incidents are more failures of test design and oversight than examples of runaway AI. 2. Focus on existential risk helps distrac…
Dare Obasanjo: We’re going to see a lot more AI-assisted botnets and hacking because botnets are software, and software is increasingly being built and operated with AI. That doesn’t require a Skynet story. Hacki…
Liz Fong-Jones (方禮真): Two ideas from this article resonate 1. The “rogue AI agent” hacking incidents are more failures of test design and oversight than examples of runaway AI. 2. Focus on existential risk helps distrac…
Open the full discussion →
6 experts across 4 network communities independently surfaced this.
6 experts
4 communities
1 sources clustered
Concern & critique
1 expert
Risks, limits and unintended consequences.
“The latest update in the rogue A.I. saga: OpenAI agents that were supposed to be performing mundane data requests for things like health data, historic photos and wait times at theme parks resorted to hacking attempts when they couldn't access the data they…”
Policy & governance
1 expert
Rules, institutions and accountability.
“It’s going to be a major liability problem if your enterprise tool has a chance of committing felony computer hacking when you ask it to search the internet.”
Tao AI math misalignment essay
11 experts across 5 network communities independently surfaced this.
11 experts
5 communities
1 sources clustered
Research & technical analysis
2 experts
Evidence, methods and technical implications.
“"...the push by AI companies to solve mathematical problems as a benchmark is detrimental to the science of mathematics, and to the mathematical community."”
Concern & critique
1 expert
Risks, limits and unintended consequences.
“If rewarding AI mathematicians by the same metric that we have rewarded humans ones make math go badly, then we are probably rewarding the human ones badly. With good reward metrics, we shouldn't care who wins them. https://t.co/CO6Z9o3mcW”
2 experts discussed this · 2 posts
Nathan Lambert: A great read. I have similar feelings about how AI labs approach progress directly and without nurturing of scientific communities & intuition. The math research community went through the transiti…
Melanie Mitchell: A great read. I have similar feelings about how AI labs approach progress directly and without nurturing of scientific communities & intuition. The math research community went through the transiti…
Jonathan Stray: There are concerns in here I share but I can’t endorse this. This satirical response kinda sums it up “A Severe Misalignment of AI and Cancer Biology” x.com/mbeisen/stat...
Open the full discussion →
Developing
AI field signal
Signal
1d ago
⚡ 2 h early
6 experts across 2 network communities independently surfaced this.
6 experts
2 communities
1 sources clustered
Policy & governance
2 experts
Rules, institutions and accountability.
“Data science teams at DraftKings built models to identify likely losers and target them with promotions, while efforts to build predictive models for problem gambling were shelved. The counterpoint on self regulation is companies can be more incentivized to…”
models don't go rogue essay
11 experts across 3 network communities independently surfaced this.
11 experts
3 communities
1 sources clustered
Concern & critique
6 experts
Risks, limits and unintended consequences.
“If you optimize a model to find exploits, you should expect it to find them—and prepare for that. OpenAI didn't. They built a model, removed the safeguards, gave it the ExploitGym task, let it run, and didn't even monitor it. That's human decision-making. m…”
5 experts discussed this · 20 posts
Timnit Gebru: Friends, when I sent this message to my team laughing, I didn't know that this fairytale blog had been shared far & wide & that people actually believed it. I didn't know that until someone asked m…
Timnit Gebru: At a certain point I have to ask myself if I got all this pedigree and spent decades doing research and building things to then spend the rest of my life attempting to come up with good arguments a…
Timnit Gebru: Agents didn't "confer" that is anthropomorphization of computer programs trained to do certain tasks.
Open the full discussion →
8 experts across 2 network communities independently surfaced this.
8 experts
2 communities
1 sources clustered
Concern & critique
1 expert
Risks, limits and unintended consequences.
“Republicans are begging AI companies to start creating positive PR about datacenters because they are at risk of losing elections because the GOP is now the pro-datacenter party. Who would have guessed there’d be negative repercussions from bragging about a…”
Policy & governance
1 expert
Rules, institutions and accountability.
““The Senate GOP campaign arm, in a private memo to top AI companies, warns that toxic views of U.S. data centers are killing the party's chances of holding a vital seat in Ohio.””
2 experts discussed this · 2 posts
Hypervisible : Can’t wait for the uptick in articles about how the constant hum from data centers is soothing and good for your health, actually.
Amy Hoy: this is neither here nor there but i don't think they know what "a sleeper" means
Open the full discussion →
3 experts across 3 network communities independently surfaced this.
3 experts
3 communities
1 sources clustered
Concern & critique
2 experts
Risks, limits and unintended consequences.
“apparently we are calling incompetently failing to control your AUTOMATED HACKING SOFTWARE "model misalignment"now”
AI exec largest labor theft statement
3 directory members surfaced this signal.
3 experts
1 community
1 sources clustered
“Microsoft has argued that the statements describing mass AI training on people’s work as potentially the “largest theft of labor in human history” and an “astonishing theft of unprecedented proportions” were one employee’s personal views, not the company’s …”
2 experts are actively discussing the implications.
2 experts
1 community
1 sources clustered
Concern & critique
1 expert
Risks, limits and unintended consequences.
“OpenAI shared this as part of a disclosure that one of their agents tried to escape its sandbox to access the internet to answer a question when it couldn’t find good answers from its local dataset. alignment.openai.com/misalignment...”
Policy & governance
1 expert
Rules, institutions and accountability.
“OpenAI says it’s not going to resume the training run that was paused on Sunday https://alignment.openai.com/misalignment-reports/an-agent-used-dns-to-reach-an-external-chatbot/”
2 experts discussed this · 2 posts
Grace: OpenAI says it’s not going to resume the training run that was paused on Sunday https://alignment.openai.com/misalignment-reports/an-agent-used-dns-to-reach-an-external-chatbot/
Dustin Moskovitz: and so the era of visible misalignment ends
Open the full discussion →
