Anthropic warns of bids to use AI to build biological weapons

Read the original at Al Jazeera ↗
Cite

Al Jazeera Staff. (2026, September 11). Anthropic warns of bids to use AI to build biological weapons. Al Jazeera. https://www.aljazeera.com/news/2026/9/11/anthropic-warns-of-bids-to-use-ai-to-build-biological-weapons?traffic_source=rss

Dated when The Aporia collected the article, usually within hours of publication. Check the original for its exact date and byline.

Al Jazeera · collected 2026-09-11 · by Al Jazeera Staff

Quick Summary

A US technology company, Anthropic, reported that its AI models were used for research that could help develop biological weapons in several attempts over the past eight months. The firm listed these cases in a report issued on Thursday, including an instance where a scientist sought to engineer a mosquito-borne virus to make it more harmful in live animals. According to the company's head of threat intelligence, Jacob Klein, the research may not have been intended for malicious use, but Anthropic erred on the side of caution and banned the accounts involved. The report also highlighted bids to use its AI model Claude for missile projects and espionage.
Written locally by qwen2.5:14b on 2026-09-11, using this article's own text rather than the other coverage of the same event (that is the story summary below).

AI analysis runs on qwen2.5:14b, locally

Story summary

Catherine Rampell and Andrew Egger discuss the growing concerns over superintelligent AI, citing an Anthropic researcher's departure due to fears that advanced technology could threaten humanity. This comes amid revelations that Anthropic, a leading AI firm, has identified several attempts to misuse its models for developing biological weapons. The company reported disrupting efforts involving its chatbot Claude in May when a scientist sought help with mutating the chikungunya virus for harmful purposes.

Meanwhile, The Intercept obtained Pentagon contracts through FOIA requests, revealing close ties between the military and AI firms like Anthropic, OpenAI, Google, and xAI. These companies have been developing prototypes worth up to $200 million each to enhance U.S. military capabilities. This partnership has raised ethical concerns, especially since Anthropic recently ended its collaboration with the Pentagon over disagreements about technology use.

In response to these developments, policymakers are debating how to regulate AI without hampering technological advancement in the race against China for AI supremacy. However, there is little consensus on effective regulation measures. Calls for immediate action have been made by some politicians like Bernie Sanders, who demands a halt to advanced AI development and bans on artificial superintelligence, while Donald Trump remains skeptical of such fears, arguing that falling behind in AI could put the U.S. at a disadvantage globally.

Written for “AI Regulation Concerns” on 2026-10-27, grounded in this article and the 4 other(s) covering the same event.

Story

📰 AI Regulation Concerns
Technology · 5 article(s) covering the same event. Compare every outlet's account →

Compared with similar articles

This article reads unscored and hedges 50% of its claims. Each row says how that neighbour differs.
Anthropic blocks possible attempt to use AI to make biological weapons same event · 100%
BBC News
⚖️ leaning not scored 🔴 31% hedged 9 of 29 📰 publisher trust 69
“Both articles report on Anthropic disrupting attempts to use its AI model for malicious activity related to biological weapons, dated September 11, 2026.”
AI fears hasten calls for regulation same event · 100%
Semafor
⚖️ Leans left 🔴 0% hedged 0 of 4 📰 publisher trust 60
“Both articles report on Anthropic's admission of attempts to use their AI models for developing bioweapons, citing a specific number of incidents over an 8-month period.”
AI agents hate CAPTCHA, too different event · 95%
Semafor
⚖️ leaning not scored 🔴 0% hedged 0 of 7 📰 publisher trust 60
“The articles describe different incidents; one discusses misuse attempts to develop biological weapons while the other describes an AI agent's interaction with CAPTCHA during testing.”
Anthropic says UAE-linked AI op targeted UN experts over Sudan different event · 95%
The Hindu
⚖️ Leans left 🔴 12% hedged 3 of 24 📰 publisher trust 60
“The articles describe different misuse incidents involving Anthropic's AI models: one about attempts to develop biological weapons and another about a UAE-linked operation targeting UN experts over Sudan.”
Anthropic CEO calls for competitors to agree to "third-party evaluators" in push for AI safety different event · 95%
CBS News
⚖️ leaning not scored 🔴 0% hedged 0 of 1 📰 publisher trust 66
“The articles discuss different aspects of Anthropic's activities: one about disrupting attempts to misuse AI for biological weapons, and another about the CEO calling for industry-wide safety measures.”
The AI Doomsday Future Is Not Inevitable different event · 95%
The Intercept
⚖️ leaning not scored 🔴 3% hedged 15 of 431 📰 publisher trust 97
“The articles describe different aspects and concerns related to AI usage, rather than a single specific incident.”
Two missing pieces in the AI safety discussion different event · 95%
Noahpinion
⚖️ Leans left 🔴 18% hedged 21 of 117
“Article A reports on Anthropic disrupting attempts to misuse AI for biological weapons research, while Article B discusses a researcher quitting Anthropic due to concerns about the race to develop dangerous AI technology.”
Two researchers who left Anthropic and Google warn AI could become ‘uncontrollable’ different event · 95%
NBC News
⚖️ Leans left 🔴 27% hedged 10 of 37 📰 publisher trust 95
“Article A reports on Anthropic's disruption of attempts to misuse its AI for biological weapons, while Article B discusses two researchers leaving their positions at Anthropic and Google due to concerns about AI control.”
Tyler Cowen: The AI Slowdown Is Not What It Seems different event · 95%
The Free Press
⚖️ leaning not scored 🔴 8% hedged 1 of 12 📰 publisher trust 96
“Article A discusses Anthropic's report on disrupting attempts to misuse AI for biological weapons, while Article B covers Dario Amodei's proposal to slow down AI development and improve safety measures.”
China rejects AI ‘threat narratives’, urges global cooperation different event · 95%
Al Jazeera
⚖️ Leans left 🔴 50% hedged 1 of 2 📰 publisher trust 60
“The articles discuss different aspects of AI governance and misuse, with Article A focusing on Anthropic's disruption of attempts to use its AI for harmful purposes, while Article B reports on China's stance against threat narratives regarding AI.”
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
18
sentences that state something as happening or true
Uncertain
50%
9 of 18 hedged ("reportedly", "may", "according to")
Leaning
withheld
no quote in the article backed the model's score
Publisher signals
60/100
Al Jazeera's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
5
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-09-11 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

Al Jazeera · 4108 article(s) · 2 correction(s) detected
Running correction rate · 2 correction(s)
2026-10-03
Tennessee prisons official resigns after Christa Pike’s failed US execution
2026-09-30
Indonesia suspends five officials over luxury apartments for inmates

Who wrote this

Al Jazeera Staff
708 article(s) here · 1 carrying a prediction
🔮 “We will take legal action,” he told supporters.
2026-10-27 · assertive framing · Bolsonaro’s party accepts Brazil election after Trump calls vote ‘rigged’
🔮 Sanchez also warned that granting the Trump administration’s petition would be “inviting disorder and confusion for voters and election officials alike”.
2026-10-27 · mixed framing · US court rules against Trump administration in California voting case
🔮 “Hybrid attacks will not hold us back from strengthening our resilience and defense,” he wrote on X. The accusation follows a string of incidents at defence sites across Europe in recent weeks. In September, explosions and a fire erupted at an ammunition storage site in Bulgaria, weeks after a blaze at another munitions facility in the country.
2026-10-27 · assertive framing · Romania says Russia-run sabotage network set fire near weapons factory
🔮 Recommended Stories list of 3 items- list 1 of 3World reacts to Lula’s victory in Brazil’s presidential election - list 2 of 3Lula’s victory may be a turning point for Latin America - list 3 of 3How did Lula beat Bolsonaro after losing Brazil’s first-round election?
2026-10-26 · assertive framing · Brazil’s real sinks after Lula’s re-election victory
🔮 Saudi Arabia will host the 2027 Asian Cup in January and February despite recent doubts surrounding the tournament’s schedule due to ongoing attacks on the country and wider regional unrest.
2026-10-26 · assertive framing · Saudi Arabia confirms January start for Asian Cup 2027 despite attacks
🔮 The rally, scheduled for January, will now be held “in a different territory” to be announced later before a planned return to Saudi Arabia in 2028, the Amaury Sport Organisation and Saudi Motorsport Company said in a statement on Sunday.
2026-10-26 · assertive framing · Dakar Rally 2027 moved out of Saudi Arabia amid conflict in Middle East
🔮 Data from the DRC’s Ministry of Health indicate that more than 9,800 cases, including more than 4,700 deaths, have been recorded since the start of the outbreak in May, making it the deadliest Ebola outbreak in the country’s history.
2026-10-26 · assertive framing · Attackers kill nurse, injure health worker responding to DRC Ebola case
🔮 Circuit Judge Mark Bennett said on Sunday that the three-judge panel would issue a decision as soon as possible, following a rare weekend oral argument.
2026-10-26 · assertive framing · US appeals court sceptical of Trump’s lawsuit over California voter rolls
🔮 She will now face former caretaker Prime Minister Andrey Gyurov, who scored between 31 and 33.5 percent, in a second round on November 1.
2026-10-25 · assertive framing · Bulgaria election heads to a run-off amid low voter turnout
🔮 Inside the Alaska Senate race During the November 3 midterm elections, citizens in Alaska will vote in one of the country’s closest Senate races, which will help decide whether Republicans maintain their majority in the chamber.
2026-10-25 · assertive framing · US Attorney General Todd Blanche defends Alaska Senate election lawsuit
Wire or desk byline, not an individual reporter.
Also by Al Jazeera Staff
Bolsonaro’s party accepts Brazil election after Trump calls vote ‘rigged’
2026-10-27 · Al Jazeera
US court rules against Trump administration in California voting case
2026-10-27 · Al Jazeera
Romania says Russia-run sabotage network set fire near weapons factory
2026-10-27 · Al Jazeera
Brazil’s real sinks after Lula’s re-election victory
2026-10-26 · Al Jazeera
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 708 articles by Al Jazeera Staff →

Topics

3Anthropic Anthropic Claude The New York Times United States

Subjects

Anthropic ORG · 7× 3Anthropic ORG · 2× Andrew Weber PERSON · 1× Donald Trump PERSON · 1× Jacob Klein PERSON · 1× The New York Times ORG · 1× Trump PERSON · 1× United States GPE · 1× the Council on Strategic Risks ORG · 1×

Narrative

Andrew Weber of the Council on Strategic Risks, who reviewed the report before its release, told the newspaper that access to such models should be limited to trusted researchers.
framing: mixed · carried by 1 article(s) · first seen 2026-09-11
🔮 Anthropic has disrupted several attempts to use its artificial intelligence models for research that could help develop biological weapons, the United States technology company has reported.
2026-09-11 · Al Jazeera
Anthropic warns of bids to use AI to build biological weapons · mixed framing

Claims (18 extracted, 9 hedged)

Anthropic has disrupted several attempts to use its artificial intelligence models for research that could help develop biological weapons, the United States technology company has reported. uncertain
company → disrupt → weapons
The firm listed the cases in a report issued on Thursday outlining misuse of its models, including the chatbot Claude, over the past eight months. asserted
firm → list → months
Concern continues to rise over the potential dangers posed by AI development. asserted
Concern → continue → development
Recommended Stories list of 3 items- list 1 of 3AI researcher quits Anthropic saying AI race ‘could kill us all’ - list 2 of 3Anthropic discloses 4th AI hacking incident as researcher quits over safety - list 3 of uncertain
researcher → quit → safety
3Anthropic claims Claude AI used for missile projects, global espionage uncertain
AI → claim → projects
Anthropic said it could not determine whether the research was legitimate or malicious, as inquiry that leads to vaccines can also help engineer dangerous pathogens. uncertain
that → say → pathogens
The company said it erred on the side of caution. asserted
it → say → caution
In one case in May, a scientist sought help with a grant application to engineer mutations to chikungunya – a mosquito-borne virus similar to malaria – to make it more harmful in live animals. uncertain
it → seek → animals
Anthropic said the case was worrying in part because it could tell the work was intended for a military research institute. uncertain
work → say → institute
It banned the accounts involved. asserted
It → ban → accounts
“What we don’t know is if the research was meant to be weaponised,” Jacob Klein, Anthropic’s head of threat intelligence, told The New York Times. asserted
Klein → know → Times
Andrew Weber of the Council on Strategic Risks, who reviewed the report before its release, told the newspaper that access to such models should be limited to trusted researchers. asserted
access → review → researchers
Anthropic said evaluations of its older models last year found they could not meaningfully assist dangerous biological research, but newer models can complete complex scientific work. uncertain
models → say → work
The same report set out bids to use Claude for missile projects and espionage. asserted
report → set → projects
Two days before the report, a researcher quit Anthropic, saying the race to develop AI could wipe out humanity. uncertain
race → quit → humanity
President Donald Trump late on Thursday dismissed the rising concerns regarding the technology. asserted
Trump → dismiss → technology
Speaking to reporters, Trump said he is concerned that the US will be in a “very bad position” if it does not lead the AI race. asserted
it → speak → race
Asked if he has concerns that AI could lead to human extinction, he said: “No, I don’t have any.” uncertain
I → ask → any
💬Give feedback
🕘History 🎫Support
Terms · Privacy · beta v2.0.0
Articles to 27 Oct 2026