A weapons-development cell based in northern Yemen allegedly used Anthropic’s Claude AI in an effort to develop guidance software for advanced rockets and missiles, according to a new report by Anthropic. The company said the actors used Claude Code in place of human software engineers, raising fresh concerns about the use of artificial intelligence in weapons development. However, Anthropic did not publicly identify the group as the Houthis, although the cell was operating in Houthi-controlled northern Yemen, according to reports. The company also said there was no evidence that the group successfully fielded an operational weapon.
Claude Used as a Virtual Engineering Team
According to Anthropic’s September report on detecting and countering AI misuse, the Yemen-based cell was working on three weapons programmes: a guided rocket, a multi-stage ballistic missile with a stated range of more than 2,000 km, and a multi-variant system referred to as the “R2000” set, which included a hypersonic glide vehicle variant. Rather than simply asking an AI model to solve a difficult engineering problem, the operators reportedly ran several Claude instances simultaneously and assigned different roles to them. Claude was used to help develop guidance, navigation and control software for the weapons.
AI Helped Develop Flight-Control Software
Anthropic said the actors used Claude to integrate an open-source autopilot system onto a phone-class flight computer. The AI was reportedly used to write control and position-estimation software, tune settings and assist with simulations. The operators also attempted to conceal the purpose of their work and divided requests across multiple sessions to avoid triggering safeguards. Anthropic said its systems blocked many requests, but not all.
Guided Rocket Test Appears to Have Failed
Anthropic said it found no evidence that the cell successfully deployed an operational missile. However, the group did conduct a test of a guided rocket, which appears to have failed. Within hours of the test, the operators reportedly returned to Claude with post-test telemetry to diagnose what went wrong.
Anthropic Bans Accounts
Anthropic said it identified the activity through its internal investigations, banned the accounts associated with the operation and shared relevant intelligence with authorities and industry partners. The company also noted that the actors had already developed an offline simulation toolkit, meaning the weapons-development effort could continue without relying on Claude or similar engineering platforms.
Also Read: Prajwal Revanna Case: Intimate Videos Found On Seized Device During Bengaluru Jail Raid
Shamshad Ali is a New Delhi–based journalist with over six years of experience spanning breaking news writing, on-the-ground reporting, and editing. When he isn’t chasing headlines, Shamshad is usually exploring new destinations, discovering hidden culinary gems, and striking up conversations with strangers. Many of whom turn out to be great sources, and even better friends. He can be reached at shamshadali.dar@gmail.com.
The post Did Yemen’s Houthis Use Anthropic’s Claude AI To design Missiles? appeared first on NewsX.