Anthropic Report Reveals State Actors Using AI for Weapons and Cyberattacks

A new report by Anthropic details how state-sponsored actors, including Russia, China, Iran, and the Houthis, utilized its Claude AI model to develop weapons, conduct cyberattacks, and spy. The findings highlight significant security risks as malicious users exploit generative AI for military and intelligence operations.

11 Sept, 11:32 — 11 Sept, 22:55

Post Share

Why this matters

This report is the first major disclosure by an AI company of state actors actively using commercial AI models for military purposes. It raises urgent questions about AI safety, export controls, and the potential for AI to lower the barrier to weapons development. The findings could accelerate global regulatory efforts similar to those for nuclear or biological weapons.

Key entitiesAnthropic Claude Russia China Iran Houthis (Yemen)Ukraine

The Story

What 16 sources agree on, dispute, and miss

What sources agree on

Anthropic released a detailed report documenting misuse of its Claude AI model by state-affiliated actors.
The report identifies Russia, China, Iran, and Houthi fighters in Yemen as users of Claude for malicious purposes.
Anthropic stated it disrupted several attempts to use Claude for weapons development and cyber operations.
The report is over 150 pages and includes case studies of AI-assisted military activities.

Key claims2 agreed · 1 disputed · 3 unverified

✗

Anthropic's report is 154 pages long

disputed·jerusalem-post

Disputed byberlingske

Berlingske describes the report as 150 pages, not 154.

✓

Houthi fighters used Claude to develop guided weapons and ballistic missiles

agreed·arab-newsexpress-tribunele-figaroklix-banaftemporiki+2

Russia used AI in cyberattacks against Ukraine, targeting WhatsApp and government ministries

unverified·kyiv-independent

Iran used Claude for military and propaganda strategies

unverified·jerusalem-post

✓

Anthropic disrupted illegal use by a Yemen-based group running three weapons development programs

agreed·express-tribunenaftemporikile-figaro

Attempts to use Claude for biological weapons research were identified and blocked

unverified·berlingsken1-serbia

Where they diverge

Primary geographic focus of the threat

kyiv-independent

Emphasizes Russian cyberattacks against Ukraine, targeting WhatsApp and government ministries.

jerusalem-post

Highlights Iranian use of AI for military and propaganda strategies.

arab-newsexpress-tribunele-figaroklix-banaftemporiki+1 moreiefimerida

Focuses on Houthi use of Claude for guided weapons and ballistic missile development.

zerohedge

Mentions both Russian and Chinese threat actors, with a broader geopolitical tone.

Characterization of Anthropic's role

Coverage gaps

Specific mention of biological weapons research attempts

Reportedberlingsken1-serbia

Missingkyiv-independentReuterszerohedge+10

Targeting of WhatsApp accounts and government ministries in Ukraine

Reportedkyiv-independent

Missingall others

Iranian use of AI for propaganda strategies

Reportedjerusalem-post

Missingkyiv-independentReuters

The Story

Analyzing sources…

Source Diversity

Excellent (92/100)

16 sources33/33

Spectrum spread— 4/5 buckets covered25/33

Far L

Left4

Left (4)

kyiv-independentn1-serbiatvn24iefimerida

Center6

Center (6)

investing-comReutersmkd-mkAPklix-baexpress-tribune

Right5

Right (5)

berlingskejerusalem-postarab-newsnaftemporikile-figaro

Far R1

Far Right (1)

zerohedge

Geographic diversity— 13 regions34/34

US3Greece2Ukraine1Intl1North Macedonia1Denmark1Serbia1Poland1Israel1Bosnia1Saudi Arabia1Pakistan1France1

Political spectrum

LeftCenterRight

kyiv-independent

n1-serbia

tvn24

iefimerida

investing-com

reuters

mkd-mk

klix-ba

express-tribune

naftemporiki

berlingske

jerusalem-post

arab-news

le-figaro

zerohedge

Outlet rating This story

SourceOutletiGeneral editorial lean of the outletStoryiHow they covered this specific storyToneFactuality

kyiv-independentcenter-leftcenter-lefturgentMostly FactualFocus on Russian aggression against Ukraine

n1-serbiacenter-leftcenter-leftcautious—Emphasizes Anthropic's blocking of misuse

tvn24

Coverage leans: balanced

Coverage spans a wide political spectrum from center-left to right, but most outlets report the facts neutrally. The overall lean is balanced, with no dominant partisan narrative.

Sources

Showing 8 of 16 sources

ReutersVery High6h ago

How Anthropic says Claude was used for weapons, spying and cyber operations - Reuters

How Anthropic says Claude was used for weapons, spying and cyber operations Reuters

Read full article →

APVery High7h ago

Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - AP News

Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says AP News

Read full article →

jerusalem-postMostly Factual8h ago

Iran used AI to develop military, propaganda strategies, Anthropic report reveals

Anthropic's 154-page report revealed the dangers of US and Western adversaries using artificial intelligence models to pose security threats.

Read full article →

investing-com6h ago

Factbox-How Anthropic says Claude was used for weapons, spying and cyber operations

By Reuters

Read full article →

kyiv-independentMostly Factual3h ago

Russia uses AI in cyberattacks against Ukraine, Europe, targeting WhatsApp accounts, government ministries, Anthropic says

"A common theme of the targeting was Ukraine and military drone technology providers and supply chains," Anthropic wrote in a report on countering AI misuse.

By Volodymyr Ivanyshyn

Read full article →

arab-newsMixed11h ago

Yemen's Houthis used Claude AI to build guided weapons - arabnews.com

Yemen's Houthis used Claude AI to build guided weapons arabnews.com

Read full article →

express-tribuneMostly Factual14h ago

Anthropic says Yemen fighters used AI to seek guided weapons

Anthropic said it disrupted illegal Claude use by a Yemen-based group running three weapons development programmes

By AFP

Read full article →

zerohedgeLow8h ago

Anthropic Says Russian, Chinese Threat Actors Used Its AI Model Claude For Malicious Activity

Anthropic Says Russian, Chinese Threat Actors Used Its AI Model Claude For Malicious Activity Authored by Aldgra Fredly via The Epoch Times, Anthropic said on Sept. 10 that it had disrupted malicious campaigns involving the use of its artificial intelligence model Claude, including operations allegedly linked to threat actors in China and Russia. The company said the threat actors include suspected state-sponsored groups, financially motivated criminals, commercial spyware ve...

By Tyler Durden

Read full article →

Coverage Timeline

First report: le-figaro · 11 Sept, 11:06|Full coverage: 16 · 12h|Window: 12h

Left-leaningCenterRight-leaning

le-figaro11 Sept, 11:06First to report

Yémen : comment des «acteurs malveillants» ont utilisé l’IA d’Anthropic pour tenter de développer des armes téléguidées

express-tribune11 Sept, 11:32

Anthropic says Yemen fighters used AI to seek guided weapons

1h later

naftemporiki11 Sept, 13:02

Anthropic: Μια ομάδα στην Υεμένη επιχείρησε να χρησιμοποιήσει το chatbot Claude για την κατασκευή κατευθυνόμενων ρουκετών και πυραύλων

1h later

arab-news11 Sept, 14:17

Yemen's Houthis used Claude AI to build guided weapons - arabnews.com

3h later

klix-ba11 Sept, 17:06

Nezgodno iznenađenje za Saudijce: Huti za pravljenje i navođenje balističkih raketa koristili umjetnu inteligenciju

jerusalem-post11 Sept, 17:17

Iran used AI to develop military, propaganda strategies, Anthropic report reveals

zerohedge11 Sept, 17:35

Anthropic Says Russian, Chinese Threat Actors Used Its AI Model Claude For Malicious Activity

1h later

AP11 Sept, 18:41

Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - AP News

iefimerida11 Sept, 19:03

Κόλαφος η έκθεση της Anthropic για την ΑΙ -Παραδέχεται ότι ένοπλοι στην Υεμένη χρησιμοποίησαν το Claude για ανάπτυξη πυραύλων

tvn2411 Sept, 19:07

Rozwój AI jak wzbogacanie uranu. Niepokojące słowa eksperta

n1-serbia11 Sept, 19:12

Antropik blokirao upotrebu AI modela za stvaranje biološkog oružja

berlingske11 Sept, 19:14

Berlingske har gennemgået rapport, som udstiller »statsstøttet« våbenudvikling med AI: Her er de mest markante nedslag

mkd-mk11 Sept, 19:23

Хутите конструирале ракети со помош на вештачка интелигенција

Reuters11 Sept, 19:25

How Anthropic says Claude was used for weapons, spying and cyber operations - Reuters

investing-com11 Sept, 19:30

Factbox-How Anthropic says Claude was used for weapons, spying and cyber operations

3h later

kyiv-independent11 Sept, 22:55Latest update

Russia uses AI in cyberattacks against Ukraine, Europe, targeting WhatsApp accounts, government ministries, Anthropic says

Story evolution

The narrative evolved from a specific Houthi weapons story to a broader state-actor threat, and finally to a call for international AI regulation.

Initial Houthi-focused reportsSept 11, 11:00-14:00 UTC

Early articles (le-figaro, express-tribune, naftemporiki, arab-news) emphasize Houthi use of Claude for guided weapons.

Broader coverage including Iran and RussiaSept 11, 17:00-23:00 UTC

Later articles (jerusalem-post, zerohedge, kyiv-independent) expand to include Iranian propaganda and Russian cyberattacks.

Regulatory and alarmist framingSept 11, 19:00-20:00 UTC

Some outlets (tvn24, iefimerida) shift to discussing AI regulation and existential risks, using the report as a case study.

Anthropic Report Reveals State Actors Using AI for Weapons and Cyberattacks

Why this matters

The Story

What sources agree on

Key claims2 agreed · 1 disputed · 3 unverified

Where they diverge

Coverage gaps

The Story

Source Diversity

Source Diversity

Political spectrum

Sources

How Anthropic says Claude was used for weapons, spying and cyber operations - Reuters

Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - AP News

Iran used AI to develop military, propaganda strategies, Anthropic report reveals

Factbox-How Anthropic says Claude was used for weapons, spying and cyber operations

Russia uses AI in cyberattacks against Ukraine, Europe, targeting WhatsApp accounts, government ministries, Anthropic says

Yemen's Houthis used Claude AI to build guided weapons - arabnews.com

Anthropic says Yemen fighters used AI to seek guided weapons

Anthropic Says Russian, Chinese Threat Actors Used Its AI Model Claude For Malicious Activity

Coverage Timeline

Yémen : comment des «acteurs malveillants» ont utilisé l’IA d’Anthropic pour tenter de développer des armes téléguidées

Anthropic says Yemen fighters used AI to seek guided weapons

Anthropic: Μια ομάδα στην Υεμένη επιχείρησε να χρησιμοποιήσει το chatbot Claude για την κατασκευή κατευθυνόμενων ρουκετών και πυραύλων

Yemen's Houthis used Claude AI to build guided weapons - arabnews.com

Nezgodno iznenađenje za Saudijce: Huti za pravljenje i navođenje balističkih raketa koristili umjetnu inteligenciju

Iran used AI to develop military, propaganda strategies, Anthropic report reveals

Anthropic Says Russian, Chinese Threat Actors Used Its AI Model Claude For Malicious Activity

Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - AP News

Κόλαφος η έκθεση της Anthropic για την ΑΙ -Παραδέχεται ότι ένοπλοι στην Υεμένη χρησιμοποίησαν το Claude για ανάπτυξη πυραύλων

Rozwój AI jak wzbogacanie uranu. Niepokojące słowa eksperta

Antropik blokirao upotrebu AI modela za stvaranje biološkog oružja

Berlingske har gennemgået rapport, som udstiller »statsstøttet« våbenudvikling med AI: Her er de mest markante nedslag

Хутите конструирале ракети со помош на вештачка интелигенција

How Anthropic says Claude was used for weapons, spying and cyber operations - Reuters

Factbox-How Anthropic says Claude was used for weapons, spying and cyber operations

Russia uses AI in cyberattacks against Ukraine, Europe, targeting WhatsApp accounts, government ministries, Anthropic says

Story evolution

Related Stories

Apple’s $164bn Rally Hinges on iPhone Duo Adoption

Apple iPhone 18 Pro pre-orders begin amid performance rumors

UK Government Rejects Calls for AI 'Kill Switch' Amid Safety Concerns

US Lawmakers Push for Stricter AI Safety Rules Following Extinction Warnings

Why this matters

Story evolution

The Story

What sources agree on

Key claims2 agreed · 1 disputed · 3 unverified

Where they diverge

Coverage gaps

Political spectrum

Coverage matrix(16 sources)