OpenZeppelin co-founder warns DeFi is unsafe due to AI

OpenZeppelin’s Manuel Araóz warns all DeFi protocols are unsafe due to AI-driven hacks, urging users to exit even top platforms as risks and losses escalate.

Manuel Araóz, co-founder of OpenZeppelin, has declared that all decentralized finance (DeFi) protocols are now unsafe. He attributes this to the rapid advancement of artificial intelligence (AI) coding agents, which have become "superhuman" at detecting vulnerabilities in smart contracts. Araóz has advised friends and family to withdraw from all DeFi positions, including major protocols like Aave, MakerDAO, and Compound. He emphasized the growing imbalance between attackers and defenders: while defenders must fix every bug, attackers only need to exploit one to steal funds. This warning follows a surge in DeFi hacks, with losses exceeding $630 million in April 2026 and over $1.1 billion lost in the past year. The rise of advanced AI models capable of autonomously finding and exploiting software flaws has heightened concerns, as DeFi’s transparent code becomes increasingly difficult to secure. Confidence in the sector has dropped, reflected by a significant decline in total value locked.