Terror groups, China turn to off-the-shelf AI to make weapons smarter
After publishing: 6 headlines quietly rewritten6 changes
- bostonherald.comseen 13 Sept, 21:48headline rewritten
Users in Houthi-held Yemen tried to develop advanced weapons with AIUsers in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - jns.orgseen 13 Sept, 21:50headline rewritten
Anthropic reveals Yemenis vibecoding ballistic missile guidance, Iran using AI to target US NavyAnthropic reveals Yemenis vibecoding ballistic-missile guidance, Iran using AI to target US Navy - nbcsandiego.comseen 13 Sept, 21:50headline rewritten
Users in Yemen tried to develop weapons with AI, Anthropic saysUsers in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - nbcdfw.comseen 13 Sept, 21:50headline rewritten
Users in Yemen tried to develop weapons with AI, Anthropic saysUsers in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - nbcconnecticut.comseen 13 Sept, 21:58headline rewritten
Users in Yemen tried to develop weapons with AI, Anthropic saysUsers in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says - Le Figaro (France)seen 14 Sept, 05:59headline rewritten
Yémen : comment des «acteurs malveillants» ont utilisé l’IA d’Anthropic pour tenter de développer des armes téléguidéesMissile, mode d’emploi ? Comment les Houthis ont tenté de perfectionner leurs armes avec l’IA d’Anthropic
What it is about
Who published it, and when
14 reports left out of the cascade - too little in common with the story to place
How it spread - 26 original reports, 4 reprints
Fri, 11 September 2026
14 reports- Times of India10:36
Anthropic revealed state actors used AI for weapons development and intelligence. A China-based actor built electronic warfare software targeting Taiwan's defenses. Yemen-based actors developed guided rockets and ballistic missiles using AI. These AI-enabled threats impact regional security and shipping lanes. Indian planners must consider these evolving state and non-state AI threats.
- Le Figaro (France)11:06edited after publishing
previously
Yémen : comment des «acteurs malveillants» ont utilisé l’IA d’Anthropic pour tenter de développer des armes téléguidéesDÉCRYPTAGE - Une «cellule d’acteurs malveillants» au Yémen a utilisé Claude pour concevoir des missiles balistiques, révèle un rapport d’Anthropic publié ce jeudi. Si leurs recherches ont apparemment échoué, cette découverte alerte sur les limites du système de protections de l’IA.
- Index.hr (Croatia)16:32
TVRTKA Anthropic objavila je da je spriječila zloupotrebu svojih AI modela za razvoj biološkog oružja i navođenih projektila od strane jemenskih hutista, naglašavajući rastuće sigurnosne rizike.
- iheart.com18:10first to report
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- iheart.com18:10
A cell of Yemen-based actors—likely the Houthis—tried to use Anthropic’s Claude AI to build ballistic missiles and guidance software.
- bostonherald.com18:44edited after publishing
previously
Users in Houthi-held Yemen tried to develop advanced weapons with AIAI is transforming warfare around the world, adding to concerns about its rapid development.
4 reprints
- mymotherlode.com11 Sept, 18:59their copy
- nbcdfw.com11 Sept, 19:46their copy
- nbcsandiego.com11 Sept, 19:46their copy
- nbcconnecticut.com11 Sept, 19:46their copy
- jns.org19:18edited after publishing
previously
Anthropic reveals Yemenis vibecoding ballistic missile guidance, Iran using AI to target US Navy“We do not have evidence the actors succeeded in fielding an operational device, but they did test-fire a guided rocket,” the AI company said.
- Google News - Top Stories19:25
How Anthropic says Claude was used for weapons, spying and cyber operations Reuters Anthropic Says Iran Used Its American AI Model to Target U.S. Navy Warships WSJ Detecting and countering misuse of AI: September 2026 Anthropic Anthropic Says It Blocked Possible Efforts to Build Biological Weapons The New York Times Users in Houthi-held Yemen tried to develop advanced weapons with AI, Anthropic says AP News
Sat, 12 September 2026
5 reports- Wirtualna Polska (Poland)05:58
Irańskie wojsko miało wykorzystywać model AI Claude do prób planowania uderzeń na amerykańskie okręty, a rosyjscy użytkownicy m.in. do budowy autonomicznego roju dronów - czytamy we wnioskach z ponad 150-stronicowego raportu firmy Anthropic, twórcy Claude'a.
- Novinky (Czechia)07:01
Společnost Anthropic popsala případ ze severního Jemenu, kde byl její chatbot Claude nasazen při vývoji několika raketových systémů. Jednu z řízených raket aktéři dokonce otestovali. Test ale zřejmě selhal – a během několika hodin se Húsíové obrátili zpět na umělou inteligenci, aby zjistili proč.
- Times of India07:07
Anthropic says a Yemen-based cell used Claude Code in place of human software engineers to write missile guidance software, running three weapons programmes at once. The AI misuse report details a guided rocket, a 2,000km ballistic missile and a hypersonic glide variant. Safeguards blocked many requests, but not all. The group test-fired a rocket, it failed, and they returned to Claude within hours. Anthropic banned the accounts.
- naharnet.com10:15
A group in Yemen attempted to use Anthropic's Claude chatbot to build guided rockets and missiles, the U.S. firm has revealed, as Iran-backed Houthi rebels are on the offensive in the country. In a report published Thursday on how it had...
- Der Spiegel (Germany)14:41
Die Huthis bedrohen die internationale Schifffahrt und die globale Energieversorgung. Um Raketen zu verbessern, hat die Gruppe offenbar Anthropic-KI genutzt. Der Konzern sagt, er habe »viele Anfragen blockiert, aber nicht alle«.
Mon, 14 September 2026
2 reports- Dnevnik.hr (Croatia)08:21
- Times of India11:11
Anthropic's new threat intelligence report details how its Claude AI was exploited between December 2025 and August 2026. Cases include a Yemeni weapons cell writing missile guidance code with Claude Code, an Iran-linked naval targeting effort, a China-linked operation profiling Uyghurs in Syria, and a Mali surveillance platform covering 25 million SIM cards. Days later, Anthropic CEO Dario Amodei called for AI development to slow down.
Tue, 15 September 2026
1 report- tportal (Croatia)03:30
Korisnici Claudea na sjeveru Jemena pokušali su uz pomoć umjetne inteligencije razviti napredne projektile, uključujući hipersoničnu letjelicu i sustav navođenja izrađen od dijelova mobilnih telefona, objavio je Anthropic. Računi su blokirani, a korisnici nisu uspjeli izraditi operativno oružje, no proveli su neuspješan test navođene rakete – nakon čega su se Claudeu obratili kako bi otkrili što je pošlo po zlu
Wed, 16 September 2026
4 reports- Euronews15:20
Anthropic's latest report on the abuse of its AI model found that a Russian-linked hacking group used Claude to automate attacks on more than 20 organisations, including Ukrainian defence ministries and drone suppliers, while separate operators ran pro-Kremlin disinformation in Africa and Moldova.
- El País (Spain)15:21
La farmacéutica aspira a “convertirse en la compañía sanitaria del mundo más impulsada por la IA” y descubrir “oportunidades científicas completamente nuevas”
- Index.hr (Croatia)17:48
SKUPINA povezana s Iranom koristila je AI model Claude za prikupljanje javnih podataka o položajima američke mornarice. Stručnjaci upozoravaju na sve češću upotrebu AI-ja u vojne svrhe.
- Channel News Asia21:58
Any war footage, even if recorded by civilians, can be used as raw material for weapons development and deployment, says this researcher.