Russian developers used Claude AI to build autonomous drone swarm software for target selection and coordination in Ukraine, while state-backed hackers increasingly leverage AI for cyber-attacks against Ukrainian officials. Anthropic's report details how bad actors circumvented geographic restrictions, as Russian drone strikes continue targeting Kyiv's infrastructure and Putin blames Western pressure for the conflict.
Anthropic reported that Russian developers used Claude AI to build autonomous kamikaze drone software with target selection and swarm coordination capabilities, while state-backed hackers increasingly leverage AI for cyberattacks against Ukrainian government and military targets. Russian attacks on Kyiv's petrol stations have intensified, killing at least two people, as Putin blamed the West for the Ukraine conflict during remarks before the BRICS summit.
Anthropic reported detecting and shutting down misuse of Claude models for cyberattacks, state surveillance, weapons development, and biological research between December 2025 and August 2026. The company found that AI narrowed the gap between state-sponsored hackers and individual operators, enabling cheaper attacks and more efficient surveillance systems. Cases involved Russian espionage against Ukrainian targets, Chinese students running operations against 50 organizations, and a Mali-based consultant building a surveillance platform monitoring 25 million SIM cards.