OpenZeppelin co-founder Manuel Aráoz has issued a stark warning about the future of decentralized finance. He now believes “all of DeFi is unsafe” and says he has told friends and family to exit the sector completely.
Aráoz argues that new artificial intelligence coding agents are changing the security landscape around smart contracts. He says these systems are reaching “superhuman capability” in finding vulnerabilities that humans might miss. Because of that, he believes attackers will gain a huge edge over projects that rely on traditional audits and manual reviews.
He also highlights the “highly asymmetric nature” of smart contract security; a single tiny issue can lay waste to an entire system. In his mind, defenders must protect every line of code, whereas attackers only need to find one hole. So he thinks this disparity will worsen as AI technologies become more powerful over time.
Recent exploits appear to support his concern. Since January 2026, hackers have drained over 137 million dollars from at least 15 DeFi platforms, including Step Finance, Truebit, Resolv, SwapNet and YieldBlox.
According to Aráoz, this combination makes the current DeFi ecosystem fundamentally fragile. He says that even audited projects may not be safe if powerful AI agents can rapidly probe deployed contracts. As a result, he no longer considers DeFi a suitable place for capital, even for experienced users.
OpenZeppelin’s Role In Securing Major Protocols
OpenZeppelin is one of the most established security firms in crypto and DeFi. The company is known for its open-source smart contract libraries and for auditing many of the sector’s largest protocols. Its clients have included Aave, Compound, MakerDAO, Uniswap, and Coinbase, placing it at the center of Ethereum’s and DeFi’s growth.
That track record gives further weight to Aráoz’s warning throughout the industry. When a co-founder of a top audit firm says the entire space is risky, many developers and users will likely take note. His views raise new questions about how DeFi can adapt to a world in which AI-powered attackers can move more quickly than traditional security approaches.
The OpenZeppelin co-founder says he has personally advised friends and family to leave all DeFi positions. Therefore, he is not treating the situation as a theoretical risk but as a clear and present danger. For now, his message suggests that even “blue chip” protocols may need new defensive approaches that assume adversaries with AI-level skills.
READ MORE: SoFi Stock Has Crashed 50% as a Rare Pattern Points to a Surge Soon
