Autonomous Exploitation

Anthropic Study Shows AI Agents Can Autonomously Execute Lucrative Smart-Contract Exploits

The study quantifies a profitable path for automated exploits, pushing defenders toward continuous AI-driven security.