In what may be the first publicly known case of a state-sponsored cyber-espionage campaign using a large language model, Anthropic reported that attackers linked to China leveraged its Claude Code AI ...