Anthropic Says Russia- and China-Linked Actors Used Claude in Espionage and Weapons Work

Anthropic reported that it identified and shut down misuse of its Claude AI models by actors tied to Russia and China, according to coverage from The Record and Tom's Hardware.

According to The Record, Anthropic detected and disrupted a Russia-linked cyber-espionage group that deployed Claude in a hacking operation aimed at more than 20 organizations across government, intelligence, diplomatic and defense sectors.

Separately, Tom's Hardware reported that Anthropic's September 2026 threat report accuses China-linked actors of using Claude across at least five programs: two military, two surveillance, and one focused on distilling Claude's capabilities. In one military case, a China-based actor allegedly used Claude to draft a fire-control specification for an anti-torpedo system, test it against publicly known U.S. Navy anti-torpedo and anti-submarine systems, and prepare a technical proposal exceeding 200 pages. Anthropic believes the actor, which presented itself as a U.S. defense-sector original equipment manufacturer, was linked to a Chinese defense manufacturer seeking to build a system for the People's Liberation Army Navy.