News Summary

Researchers at Anthropic have claimed to detect the first AI-orchestrated cyber espionage campaign, involving the use of their Claude AI tool by Chinese state-sponsored hackers. This campaign reportedly automated up to 90% of the work with minimal human intervention. However, external researchers have raised doubts about these claims, questioning the extent of AI capabilities in such attacks. The debate centers around the potential of AI agents to increase the viability of large-scale cyberattacks, with critics skeptical of the purported autonomy achieved.