Iran-backed Houthi rebels in Yemen have attempted to use advanced artificial intelligence technology to develop software for ballistic missile guidance systems, according to a report released by the AI company Anthropic. The firm identified a “cell of threat actors based in northern Yemen” engaged in three weapons development programs, including efforts to build a guided rocket that employs a standard mobile phone-class flight computer enhanced by final-phase homing guidance.

Anthropic’s report highlights the group’s use of Claude, the company’s sophisticated AI model, to assist in the creation of software intended to improve missile accuracy and control. While the company did not explicitly name the Houthis, the description of the actors matches the known location and capabilities of the Iran-backed rebel faction operating in northern Yemen.

The attempted integration of AI technology into ballistic missile development marks a notable development in the evolving tactics of armed non-state groups. Typically reliant on more rudimentary or improvised guidance systems, the Houthis’ use of AI indicates an effort to enhance the precision and lethality of their missile arsenal. The guided rockets referenced in the report are designed to use commercially available flight computers, which are then augmented with AI-driven homing capabilities to potentially evade interception and improve targeting accuracy.

Yemen’s conflict, ongoing since 2014, involves multiple factions and has seen significant foreign involvement, including support for the Houthis from Iran. The rebels have launched numerous missile attacks into Saudi Arabia and other neighboring countries, often targeting civilian and military infrastructure.

Anthropic’s findings raise concerns about the proliferation of advanced AI tools in the hands of militant groups seeking to develop increasingly sophisticated weaponry. The company’s report underscores the challenges of monitoring and preventing the misuse of AI technologies designed for peaceful applications but susceptible to adaptation for military purposes by hostile actors.

The report comes amid growing global attention to the intersection of emerging technologies and security threats. Governments and international organizations are increasingly focused on establishing frameworks to regulate AI development and prevent its employment in destabilizing activities. However, the ease of access to commercially available AI systems and consumer-grade hardware complicates efforts to stem the use of these technologies by groups like the Houthis.

While the full extent of the Houthis’ AI-assisted missile program remains unclear, Anthropic’s disclosure highlights the rapid evolution of warfare dynamics in Yemen and the broader Middle East. Continued monitoring and international cooperation will be critical to addressing the risks posed by the fusion of AI and missile technology in conflict zones.