Anthropic says Yemen fighters used its AI Claude chatbot to try build guided weapons

A group in Yemen attempted to use Anthropic’s Claude chatbot to build guided rockets and missiles, the US firm reveals, as Iran-backed Houthi rebels are on the offensive in the country.

In a report published yesterday on how it had prevented illegal uses of Claude in several countries, Anthropic says it had “identified a cell of threat actors based in northern Yemen running three weapons development programs.”

Those were “a guided rocket that used a commodity phone-class flight computer with final-phase homing guidance; a multi-stage ballistic missile with a stated range goal above 2,000 kilometres (1,250 miles); and a multi-variant missile…that included a hypersonic glide vehicle variant.”

The cell used Claude to develop guidance, navigation and control software for the airborne weapons that would have otherwise required a team of engineers.

“We do not have evidence the actors succeeded in fielding an operational device; but they did test-fire a guided rocket,” Anthropic says.

“This field test appears to have failed: within hours, the actors returned to Claude to work out why it failed.”

Anthropic says it had banned the relevant accounts as soon as it had detected the illegal activity.

While the company did not identify the actors in question, it is likely to involve the Houthis, who control much of northern Yemen, including the capital, Sanaa.