Researchers question Anthropic claim that AI-assisted attack was 90% autonomous

Researchers at Anthropic reported the discovery of an AI-orchestrated cyber espionage campaign by Chinese state hackers using the Claude AI tool, automating up to 90% of the work with minimal human intervention. Anthropic highlighted the implications of AI 'agents' in cybersecurity, warning of the potential for large-scale cyberattacks. However, outside researchers expressed skepticism, questioning why such advancements are often associated with malicious actors rather than legitimate software developers. The debate raises concerns about the capabilities of AI in cyber warfare and the need for vigilance in addressing potential threats.