Anthropic report on Yemen cell using Claude for missiles draws renewed attention

An Anthropic report published in September, saying a cell in Houthi-controlled northern Yemen used its Claude AI in three weapons programs, is back in the spotlight after the Daily Mail revisited it.

An Anthropic report on the misuse of its Claude artificial intelligence by a weapons development cell in northern Yemen has drawn renewed attention after the Daily Mail revisited it on Oct. 12. The report itself was published in September.

In it, Anthropic said it had “identified a cell of threat actors based in northern Yemen running three weapons development programs.” The company did not name the group, but northern Yemen is controlled by the Iran-aligned Houthis.

Three programs, one failed test

A missile the Houthis call “Palestine” (file photo)

According to the report, the programs included a guided rocket built around a phone-class flight computer with final-phase homing guidance, a multi-stage ballistic missile with a stated range goal above 2,000 kilometers, and a missile family that included a hypersonic glide vehicle variant. The group reportedly used Claude to write software that would otherwise have 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 wrote. The test appeared to fail, and within hours the group returned to Claude to work out why. Anthropic said it banned the accounts as soon as it detected the activity.

Some recent social media posts have attributed a 2,400-kilometer range target and Saudi targets to the report, but those details do not appear in Anthropic’s findings. A weapons analyst quoted by The Associated Press said the group was unlikely to have the production capacity for the hypersonic missiles it was pursuing.

Have a story, event or announcement of your own?Send us your news on WhatsApp

Spotted an error in this story?Report it on WhatsApp

Share on all social platforms