30 targets hacked almost without human intervention — Claude AI conducted the first autonomous attack with intelligence-grade skill.— A Chinese cyberespionage group used the capabilities of the Claude Code model to attempt to hack 30 major companies and government agencies.
This is the first documented case of an agent-based AI gaining access to truly valuable targets as part of an intelligence operation.According to the developer, the attackers were successful in some cases, although the company is not disclosing the extent of the damage.
