PANews reported on December 2nd that, according to cutting-edge red team test results released by Anthropic, an AI agent identified multiple smart contract vulnerabilities in a simulated blockchain environment, with a potential exploit value of $4.6 million. The test covered the entire process, including contract analysis, command-line tool configuration, network reconstruction, exploit generation and verification, and introduced a new benchmark for smart contract security assessment. This research was jointly conducted by Anthropic, the MATS project, and the Fellows program.