[Researchers Warn: AI is Shortening the Validity of Security Audits, Protocols Require Continuous Review] According to Cointelegraph, blockchain security experts have warned that as AI tools enable hackers to identify vulnerabilities more quickly, crypto protocols need to conduct continuous reviews of smart contracts rather than relying on one-time audits. TRM Labs Policy Head Ari Redbord stated, 'The pace of attack techniques is outpacing the coverage of one-time audits.' A CertiK report revealed that hackers stole $1.32 billion in the first half of 2026 and have begun using AI tools to target the codebases of closed DeFi protocols. In June, Aztec Connect was exploited for $2.1 million three years after shutting down, and mySwap was hacked for $300,000 six months after closure.