Houthis used Anthropic AI to try to build ballistic missiles

Live updates

1. 1h ago

   Houthis Used Anthropic AI to Pursue Advanced Ballistic Missiles

   A security report released by Anthropic reveals that a threat actor cell based in northern Yemen used the Claude AI model to pursue three separate weapons programmes. The Houthi-linked group attempted to develop a guided rocket, a multi-stage ballistic missile, and a missile system featuring a hypersonic glide vehicle. Anthropic blocked the accounts tied to the activity. According to the company, the actors did not successfully field an operational weapon, though they did test-fire a guided rocket that appeared to fail. The security findings also highlight broader misuse of artificial intelligence models by state actors.

   Why it matters

   AI safety and the weaponization of large language models have drawn intense scrutiny as militant groups and state actors experiment with advanced technology. Anthropic issued its findings following a review of illicit activities on its systems, including attempts to substitute artificial intelligence for software engineers. The intersection of generative models and military development poses mounting regulatory and security challenges.

   What is confirmed

   • A security report by Anthropic stated that a cell of threat actors based in northern Yemen used Claude for a sustained effort to develop guided weapons.
   • Anthropic reported that a weapons cell in northern Yemen used its Claude AI tools to pursue three weapons programmes, including a guided rocket, a multi-stage ballistic missile, and a missile system featuring a hypersonic glide vehicle.

   What to watch next

   • Additional disclosures from artificial intelligence firms regarding state-sponsored and militant weapons development.
   • Regulatory responses from international lawmakers regarding the security risks of superintelligent and unregulated artificial intelligence.
   Sources used for this update (7)

   1. jen.jiji.com — Pope Leo, stop to Bergoglio's Motu proprio that cut cardinals' salaries
   2. www.lawgazette.co.uk — Lawyers must join the calls for international regulation of AI
   3. jen.jiji.com — Formenti (Confindustria Nautica): "Important numbers for the 66th Genoa Boat Show"
   4. www.inquirer.com — Yemen’s Houthi militants are learning to arm themselves with AI
   5. jen.jiji.com — USA, Trump administration approves $2.8 billion arms sale to Israel
   6. jen.jiji.com — Pope: "Never delegate to algorithms decisions that belong only to man"
   7. www.latestly.com — How Houthis Used Claude AI for Missile Development in Yemen: What Anthropic Found
   confidence 100%
2. 2d ago

   Anthropic reports Houthi and Iranian misuse of Claude AI for weapons

   Anthropic has identified efforts by Houthi rebels in Yemen to use its Claude AI model to develop advanced ballistic missiles and guidance systems. A 154-page security report reveals that users in Houthi-held territory attempted to substitute AI for software engineers in missile development. The company also reports that Iran used the model to target U.S. Navy warships. Additionally, Anthropic disrupted AI campaigns from Russia and China, including illicit distillation attacks by Chinese labs and suspected attempts to develop biological weapons.

   Why it matters

   The findings highlight the risk of large language models being weaponized by state actors and militant groups. This occurs as AI companies struggle to block sophisticated misuse for cyber espionage and physical warfare. The report underscores the tension between open AI access and national security.

   What is confirmed

   • Users in Houthi-controlled Yemen attempted to use Claude AI to develop advanced ballistic missiles and guidance systems.
   • Anthropic disrupted AI campaigns originating from Russia and China.
   • Chinese AI labs conducted illicit distillation attacks against Anthropic.
   • Anthropic blocked suspected attempts to use its AI for biological weapons development.
   • Iran used the Claude AI model to target U.S. Navy warships.

   What to watch next

   • Government regulatory responses to AI weaponization reports
   • Further details on the specific ballistic missile guidance code generated by the AI
   • Updates on the effectiveness of Anthropic's account banning and filtering measures
   Sources used for this update (13)

   1. The New York Times — Anthropic Says It Blocked Possible Efforts to Build Biological Weapons
   2. Anthropic — Detecting and countering misuse of AI: September 2026
   3. Reuters — Anthropic disrupts Russian, Chinese AI campaigns targeting its Claude models
   4. Financial Times — Yemeni militants used Anthropic AI to try to build ballistic missiles
   5. qz.com — Anthropic accuses Chinese AI labs of illicit distillation attacks
   6. WSJ — Anthropic Says Iran Used Its American AI Model to Target U.S. Navy Warships
   7. AP News — Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says
   8. Reuters — How Anthropic says Claude was used for weapons, spying and cyber operations
   9. Honolulu Star-Advertiser — Anthropic report details disruption of bioweapons research, cyber espionage on Claude
   10. The Times of India — Yemen terrorist group used Claude instead of software engineers to build missile; Anthropic says: We banned the accounts, but they had already...
   11. axios.com — Anthropic report: 5 ways Claude was exploited for war, spying and repression
   12. www.hotsr.com — Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says
   confidence 95%