Ethereum Foundation: AI Finds Bugs, But Humans Must Validate Them
The Ethereum Foundation’s Protocol Security team has identified a critical bottleneck in AI-assisted security research: validating findings now consumes far more time than generating them. In a July 9 disclosure, the Foundation revealed that when multiple coordinated AI agents audited systems Ethereum depends on—including peer-to-peer networking code, cryptographic libraries, and smart contracts—they produced approximately 1,000…