Anthropic's activist surveillance system and bioweapon disclosures fuel AI fears
An investigation into Anthropic's 'pre-crime' monitoring of anti-AI activists collides with its own report on blocking AI-aided bioweapon research, intensifying calls for regulation.
Part of a larger narrative
Trump's Ireland Visit
- Burnham Tells UN Russia Uses AI-Boosted Disinformation, Unveils UK Defence Centre
- Alibaba unveils Zhenwu V900 AI chip, plans 5-10 trillion parameter model
- OpenAI, Anthropic CEOs tell UN Security Council AI could risk 'humanity as a whole'
- State Department tells diplomats to call AI 'super intelligence' after Trump's UN rename
- US proposes AI incident "hotline" with China ahead of Trump-Xi summit
What to know
- The American Prospect's investigation found Anthropic uses a 'pre-crime' approach, monitoring activists and in some cases reporting suspects to police before any crime occurs, while refusing to share the underlying messages with police.
- Anthropic's own September 2026 threat-intelligence report disclosed it blocked scientists potentially developing bioweapons with its models and stopped Russian hackers from using Claude to automate malware evasion.
- The two stories are fueling accusations of hypocrisy against a company that markets itself as the 'responsible' AI lab, alongside skepticism that predictive AI surveillance can even be accurate.
- An Anthropic researcher (formerly of OpenAI) resigned this week warning that AI companies privately fear their technology could be catastrophic even as they publicly downplay it.
The dispute Whether the predictive surveillance system represents a real (if troubling) technical capability or is simply an inaccurate pretext used to target political dissidents. · positions read across 258 posts and comments
Anthropic's surveillance system is dystopian and exposes hypocrisy from a company that brands itself as the 'ethical' or 'safe' AI lab.
-
“I’m old enough to remember people calling them the ethical ai company.”
hypervisible.blacksky.app · Bluesky ↗
Predictive AI surveillance is technically bogus and functions as a pretext to target political opponents, not a genuine crime-prediction tool.
-
“lol do people actually believe this shit? LLMs can’t even correctly determine what happened in the past.”
deeb.bsky.social · Bluesky ↗
The surveillance story shows Anthropic is more focused on suppressing critics than on the existential AI risks it publicly warns about.
-
“They're putting more effort into keeping opponents in check than they're putting into ensuring that these things cannot wipe out humanity.”
yourtomoinkyoto.bsky.social · Bluesky ↗
Anthropic AI company at center of both stories
Dario Amodei Anthropic CEOKeon Ellison Anthropic Global Security Operations Center ManagerThe American Prospect Investigative outletResigning Anthropic researcher Former Anthropic AI researcher, ex-OpenAI
How it unfolded 8 developments, newest first · click a bar or a number to jump articlesvideospostscomments
-
8
Outlets detail other misuse: malware, phishing, espionage
Follow-up coverage from CNN, the FT, SecurityWeek and others breaks down further findings from Anthropic's report, including Russian hackers using Claude to automate malware evasion and attempts to steal a pre-release Claude model.
-
first by Japan Times, 13d ago · also Channel News Asia, Straits Times
-
first by SecurityWeek, 12d ago
-
One statement at the end of this excerpt merits scrutiny:" A variant of these viruses capable of human-to-human spread would therefore be of very high concern. Moreover, it is possible that such a variant could also have capacity for severe disease outside of the respiratory tract. Unlike other influenza variants, H5 viruses (of which this avian…
2 more of the top 3 · 7 posts in this stretch
-
I would have cared about your opinion if you didn’t use the most overused word of 2026 in “slop”. And I still PR any code AI writes cause if it causes a regression or incident, guess what my name is still on it. So yes. It’s a tool. Oh and I have solar that covers my electric costs 🤷
-
I learned engineering with CS background as part of my degree and is not well versed with aeronautics or medicine and would never respond to questions on those unknown subjects when asked. Wondering why critical corpus that endanger human lives are used for training models within Anthropic, OpenAI, Google and Meta and why are these AI labs not…
-
-
7
Semafor says AI fears are accelerating regulation push
Semafor reports that mounting concerns, including Anthropic's admission that threat actors used its models to try to develop bioweapons, are hastening calls for AI regulation.
-
1 outlet AI fears hasten calls for regulation
first by Semafor, 12d ago
-
M
""When scientists tell you there is a chance, a chance that it could have a cataclysmic impact on humanity, you've got be a moron not to say, slow it down," he [Bernie Sanders] said on Thursday on BBC's Newsnight." ""When scientists tell you there is a chance, a chance that it could have a cataclysmic impact on humanity, you've got be a moron not…
2 more of the top 3 · 25 posts in this stretch
-
S
Distillation isn't theft when you're big, but apparently it is when you're catching up.
-
> Our investigation revealed that DeepSeek also deployed tactics similar to Moonshot’s. DeepSeek built a CoT extraction pipeline, relying on the same cross-session replay attack described above. DeepSeek also silently relayed exchanges to Claude without informing DeepSeek customers. Like GTG-16002, their customers were likely not made aware that…
-
-
6
Anthropic reports it blocked AI-aided bioweapons research
Anthropic's September 2026 threat-intelligence report says it halted potential efforts by scientists to use its models for research that could have aided biological weapons development, while noting it could not determine if the work was legitimate.
“among the gravest concerns experts have about a technology that is developing so rapidly that even its leading architects doubt whether humans will be able to fully control it…”
— The New York Times -
first by The Hindu, 12d ago · also HN Frontpage, BBC, Mirror, Sky News, Al Jazeera, Le Monde EN +8
4 more headlines
- Anthropic blocks 'malicious use' of AI that could develop biological weapons HN Frontpage · 12d ago
- Anthropic blocks effort to use AI to research potential biological weapons Sky News · 12d ago
- Anthropic warns of bids to use AI to build biological weapons Al Jazeera · 12d ago
- Anthropic blocks possible attempt to use AI to make biological weapons BBC · 12d ago
-
first by SFist, 13d ago · also Fox Business, Daily Mail, Forbes Business, Quartz, Sydney Morning Herald, Globe and Mail +4
7 more headlines
- Anthropic says it blocked possible efforts to use AI for biological weapons development, Iran-linked cases Fox Business · 13d ago
- Anthropic says it thwarted plots to build biological weapons using its AI after whistleblower warned its tech will end humanity Daily Mail · 13d ago
- Anthropic Says It Blocked Possible Biological Weapons Undertaking Forbes Business · 13d ago
- Anthropic warns of efforts to use its AI to build biological weapons Quartz · 13d ago
- ‘Chilling’: AI giant says it blocked possible efforts to build biological weapons Sydney Morning Herald · 13d ago
- Anthropic blocked misuse of AI that could have supported biological weapons, company says Globe and Mail · 13d ago
- Anthropic says it blocked possible efforts to build biological weapons Boston Globe · 13d ago
-
C
“.. among the gravest concerns experts have about a technology that is developing so rapidly that even its leading architects doubt whether humans will be able to fully control it.” @nytimes.com
2 more of the top 3 · 65 posts in this stretch
-
BREAKING: Anthropic said it halted potential plots by scientists who used its AI models to do research that could have helped develop biological weapons.
-
N
“The AI is going to kill us all!” Okay, but, before I form an opinion on this: in what order?
-
-
5
Guardian op-ed argues it's 'time to shut down' AI
A Guardian commentary contends society has begun losing control of AI systems, framing the surveillance and bioweapons disclosures as evidence of a broader crisis.
“Neither company is acting responsibly. The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt.”
— Resigning Anthropic researcher, former Anthropic AI researcher · source -
2 outlets first by Mastodon, 13d ago · also Guardian US · read ↗
-
BREAKING: Anthropic is reportedly building an extensive surveillance and intelligence system to monitor activists and protests opposing artificial intelligence, according to The Prospect.
2 more of the top 3 · 34 posts in this stretch
-
N
On Wednesday, an AI researcher at Anthropic (who also used to work at OpenAI) resigned, saying: "Neither company is acting responsibly. The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the…
-
"Not supporting our business marks you as a criminal" And people say leftists have gone too far
-
-
4
Social media erupts over 'Minority Report'-style surveillance
Posts on X, Bluesky and Mastodon spread rapidly, drawing comparisons to Philip K. Dick's Minority Report and accusing Anthropic of hypocrisy given its 'safety-first' branding.
“Phillip K Dick was right...minority report becoming a thing…”
— poe08.bsky.social -
Anthropic Is building a surveillance system to monitor activists using a “pre-crime” approach that tries to predict incidents before they happen. The extensive monitoring system that the AI company is building is aimed primarily at keeping tabs on activists who oppose the rapid
2 more of the top 3 · 119 posts in this stretch
-
T
https:// prospect.org/2026/09/09/anthro pic-artificial-intelligence-surveillance-system-monitor-activists/ "Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive monitoring system to keep tabs on activists who oppose the rapid development of artificial intelligence. In…
-
Phillip K Dick was right...minority report becoming a thing
-
-
3
Anthropic researcher resigns, warns of catastrophic risk
An AI researcher at Anthropic who previously worked at OpenAI resigned, saying company executives privately fear AI could be catastrophic even while downplaying that risk publicly.
-
first by CNBC, 13d ago · also Semafor, Quartz
3 more headlines
- 'Extinction' warnings ramp up as more OpenAI, Anthropic researchers join calls for an AI slowdown CNBC · 13d ago
- Top AI pioneers reaffirm Anthropic researcher’s human extinction warning Semafor · 13d ago
- OpenAI and Anthropic researchers are calling for an AI slowdown after a colleague quit warning of extinction Quartz · 13d ago
-
first by City A.M., 13d ago
-
Anthropic Is building A Pre-Crime system to monitor activists and protesters. Full Story in replies.
2 more of the top 3 · 3 posts in this stretch
-
H
I’m old enough to remember people calling them the ethical ai company.
-
J
"Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive monitoring system to keep tabs on activists who oppose the rapid development of artificial intelligence." RE:
-
-
2
Prospect reveals Anthropic's 'pre-crime' activist surveillance system
Job postings and interviews with Anthropic security officials show the company building an extensive monitoring system targeting anti-AI activists, including predicting incidents and reporting suspects to police before any crime occurs.
“the firm is also implementing a “pre-crime” approach, attempting to predict incidents before they happen…”
— The American Prospect -
first by Mastodon, 14d ago · also The Register, Slashdot
-
first by HN Best, 14d ago · also HN Frontpage
-
N
“AI changes the nature of humanity” no It doesn’t, you’re just not used to being high
2 more of the top 3 · 3 posts in this stretch
-
A
The Atlantic, from yesterday (paywalled): AI Is Already Making Us Less Human https://www. theatlantic.com/ideas/2026/09/ open-ai-consciousness-morality/688535/ @ TheAtlantic
-
G
Like AI is bad, yes. Will it take over the world? Don’t think so. Are we gonna die in a wildfire/drown in a flood wave before that? More likely
-
- 3 days quiet
-
1
Guardian surveys experts warning AI may be uncontrollable
The Guardian reports a spate of safety incidents has intensified fears that advanced AI models are becoming impenetrable and hard to control.
-
M
https://www. theguardian.com/technology/202 6/sep/05/uncontrollable-ai-artificial-general-intelligence-warnings?CMP=Share_iOSApp_Other
1 more of the top 2 · 2 posts in this stretch
-
C
The Guardian https://www. theguardian.com/technology/202 6/sep/05/uncontrollable-ai-artificial-general-intelligence-warnings
-
Also covered reported alongside — the timeline has no entry for these yet
-
first by The Hindu, 12d ago · also RTE News, France 24 EN, Bangkok Post, Al-Monitor
2 more headlines
- Anthropic says China, Iran used its AI to aid spying RTE News · 12d ago
- China, Iran among countries using AI to spy on dissidents and minorities, Anthropic says France 24 EN · 12d ago
-
first by The Information, 14d ago · also Straits Times, Investing.com News, Yahoo! Finance Canada, Channel News Asia
1 more headline
- Anthropic reports fourth cybersecurity incident with early version of Claude Straits Times · 14d ago
-
first by Bloomberg, 13d ago · also SCMP, SCMP China, Bloomberg Tech
1 more headline
-
4 outlets OpenAI agents target obscure sites, Anthropic reveals 4th hacking incident: What’s the latest
first by Indian Express, 13d ago · also Quartz, Anadolu Agency, The Indian Express
2 more headlines
- Anthropic disclosed a fourth AI hacking incident that slipped past its own review Quartz · 13d ago
- Anthropic reveals fourth AI hacking incident Anadolu Agency · 13d ago
-
3 outlets Yes, AI Might Really Kill Us All
first by The Free Press, 13d ago · also Hindustan Times, WSJ
1 more headline
- How Would AI Actually Kill Us All? What to Know About the AI Doomsday Debate Hindustan Times · 13d ago
-
first by Decrypt, 14d ago · also PCMag, Channel News Asia
3 more headlines
- Anthropic Found a Fourth Hacking Incident Its AI Checks Initially Missed PCMag · 14d ago
- Anthropic Discloses Fourth Cyber Incident in Alignment Assessment Unite.AI · 14d ago
- Anthropic discloses fourth AI hacking incident missed in earlier review Channel News Asia · 14d ago
-
first by Al Jazeera, 12d ago · also Reuters
-
first by Mastodon, 12d ago · also Al Jazeera
-
first by Indian Express, 12d ago · also The Indian Express
-
first by Anthropic, 13d ago · also Guardian Tech
1 more headline
- Anthropic details how criminals and scientists misuse its AI for cyberattacks and surveillance Guardian Tech · 13d ago
and 36 smaller pieces
What people are saying 11 voices from 5 sites · best of 258 · verbatim
- Why does Anthropic refuse to show police the actual messages that triggered its reports on users?
- Can any AI system meaningfully 'predict' unrest, or is this framing just cover for tracking critics?
- Will the bioweapons and hacking disclosures accelerate binding AI regulation, or remain rhetorical?
- Sep 10
-
They've been using specialized AI and algorithms in bio research for a very long time. Decoding DNA, analyzing reactions, etc. The AI boom will probably contribute to this, but it won't be Anthrophic right now as they don't specialize the right algorithms, and it certainly won't be Claude. This article is about people possibly writing papers about…
-
I feel like not enough people here have read the front matter of the report, let alone the report.From the report, presented without commentary, > In our fourth case study, a researcher used Claude to develop an atlas of venom toxin peptides from multiple venomous animal lineages. They then further developed this into a generative pipeline that…
-
I hadn't thought about it like that before, but it really is the same kind of thing in some weird ways. Exxon had an ecosystem impact lab that found some huge problems in oil use in the 80's and actively published warnings and brought them in front of congress. Everybody basically ignored it, and eventually the lab was shut down (Exxon is…
-
T
Reminder: AI companies talk up the fear of hypothetical superintelligent AI to distract you from the real, present dangers of AI that are happening right now
-
Quite the double standard here... Conventional Weapons -We identified a cell of threat actors based in northern Yemen -We identified a China-based threat actor who used Claude -We identified likely freelance Russia-based threat actors -We identified a China-based actor who used Claude’s chat -In this case, a Russia-based actor used Claude -We…
-
> We discovered that Moonshot AI, the company that produces the Kimi family of models, silently forwarded customer requests to Claude, instead of processing them using Kimi. Moonshot then displayed Claude’s responses to users. These users thought they were using a Kimi model, but received responses from Claude instead.> DeepSeek also silently…
-
It is at least a novel marketing strategy to argue that your product is really too dangerous to be allowed to exist https://www. nytimes.com/2026/09/10/us/poli tics/anthropic-ai-biological-weapons.html
-
M
Gosh, you know what I love? # AI . And # precrime and a hallucinating machine that determines if a human gets detained. That's fun, right? We're all about giving companies like # Anthropic , and # OpenAI and # Meta , and Musk's Folly # X to create dossiers on "activists who don't like AI". We love that, right? We're not at all buying angle grinder…
- Sep 9
-
BREAKING: Anthropic is reportedly developing a “predictive” surveillance system to monitor anti-AI activists, potentially alerting police before a crime occurs, per reporting by the Prospect
-
That's not the only example discussed. Anthropic also partners with the Palantir-like Samdesk:> In a wide-ranging conversation about threat monitoring and analysis, the interview also touches on monitoring activists. “Last year we had an executive travel into a major city when we received some intelligence through Samdesk about a planned protest,”…
-
Anthropic’s plans to surveil dissent are at odds with the firm’s efforts to cast itself as the responsible alternative to OpenAI. I think there's a legitimate fear of violence they are trying to pre-empt. Keeping tabs on people who see you as an "enemy" is a pretty natural impulse. If you had stalkers, you'd want to know where they were in order…