Manuel Aráoz says he's telling friends and family to pull money out of Aave, MakerDAO, and Compound, citing an unwinnable security arms race amplified by AI.
The co-founder of one of crypto’s most trusted security firms just told everyone he knows to get out of DeFi. That’s not a random anon on Twitter. That’s the guy whose company literally writes the code libraries that most DeFi protocols are built on.
Manuel Aráoz, co-founder of OpenZeppelin, declared on May 26 that he now considers the entire DeFi sector unsafe. He went further, saying he has personally urged friends and family to withdraw their positions from major lending protocols including Aave, MakerDAO, and Compound.
The asymmetry problem #
Aráoz’s argument boils down to a concept that security professionals have understood for decades but rarely state this bluntly in public: defenders have to be right every single time, while attackers only need to be right once.
In English: a smart contract audit might catch 99 out of 100 vulnerabilities. The one it misses is all an attacker needs to drain the entire protocol. And that math was already unfavorable before AI entered the picture.
The rise of advanced AI coding agents has fundamentally shifted the landscape. These tools can scan massive codebases and identify weaknesses at a speed and scale that no human audit team can match. Attackers now have access to the same AI capabilities as defenders, but the economics still favor offense.
April 2026 saw over $600 million lost to DeFi exploits, a staggering monthly toll that has intensified scrutiny on systemic risks across the ecosystem. When protocols are deeply interconnected, sharing liquidity pools and composable smart contracts, a single exploit can cascade through the entire system.
The irony is thick enough to cut with a knife. OpenZeppelin’s own contract libraries are among the most widely used building blocks in DeFi. The company’s auditing services are considered industry-standard. OpenZeppelin even launched AI-powered security tools as recently as May 2026, designed to bolster its defenses against exactly the kind of threats Aráoz is now warning about.
And yet, the founder of that very company is saying: it’s not enough.
Why this warning hits different #
Aráoz isn’t a competing protocol trying to poach users. He’s not a regulator with an ax to grind. He runs the company that profits directly from DeFi’s continued existence and growth. When someone with that much skin in the game says the entire sector is unsafe, it’s worth paying attention.
The social media response has been predictably intense, with the crypto community split between those who view the comments as a necessary wake-up call and those who see them as irresponsible fear-mongering from someone whose company should be solving the problem, not declaring it unsolvable.
The discussion has surfaced a deeper question that the industry has been dancing around: can current auditing methodologies, even enhanced with AI, ever provide meaningful security guarantees for systems that hold billions of dollars in user funds?
Traditional software companies ship patches and updates continuously. A bug in a smart contract governing a $10B lending pool is a potential extinction event. And unlike traditional software, immutable smart contracts can’t always be patched after deployment without introducing new governance risks.
What this means for investors #
As of May 27, markets have not shown an immediate reaction to Aráoz’s statements. Token prices for the protocols he named, including Aave, MakerDAO, and Compound, haven’t seen significant movement directly attributable to his warning.
For anyone with significant capital deployed in DeFi lending protocols, this is a moment to reassess risk allocation. The $600 million lost in April alone represents a data point that’s hard to dismiss, and the trend line isn’t improving despite years of investment in better auditing tools and practices. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our