Prominent crypto security figure and OpenZeppelin co-founder warns that AI-powered coding agents have made DeFi unsafe and advises friends and family to exit DeFi positions, including blue-chip protocols like Aave, MakerDAO, and Compound.

He states that all of DeFi is unsafe, signaling heightened concern about security risk across the DeFi ecosystem.

The warning targets core DeFi platforms—Aave, MakerDAO, and Compound—that OpenZeppelin has helped secure since 2015.

The market reaction is cautious, with investors considering diversification and self-custody, though there has not been a mass withdrawal yet.

The debate centers on immutable on-chain contracts, the possibility of AI-driven exploits, and broader implications for tokenized assets and institutional participation.

April exploits totaled hundreds of millions, including the Drift and Kelp DAO attacks, linked to North Korean-linked hackers, contributing to a surge in reported exploits.

May has seen ongoing DeFi security incidents, such as Verus Network’s breach and a Polymarket security incident, underscoring persistent security concerns.

Industry pushback argues that many 2023-2024 losses stem from operational security and misconfigurations rather than pure code flaws.

DeFi’s growth brings billions locked in, emphasizing that risk may be intrinsic to the ecosystem rather than confined to individual platforms.

Some acknowledge risk but note AI could also bolster defense through stress testing and red-teaming of smart contracts.

Overall, the warning renews questions about DeFi’s viability, safety standards, and regulatory considerations, while highlighting ongoing efforts in formal verification, audits, and risk mitigation.

The warning has amplified on social platforms, influencing sentiment and drawing broad attention.