For years, the decentralized finance (DeFi) sector has prided itself on being a more transparent and resilient alternative to traditional banking. But now, a figure widely regarded as a founding voice in crypto security has dropped a bombshell: they no longer consider any DeFi protocol safe to use. The statement, made during an interview earlier this week, sent shockwaves through an industry already reeling from billions in losses due to hacks and exploits.
The Warning
“I now consider all of DeFi unsafe,” said the pioneer, who asked not to be named due to ongoing security concerns. The expert, who has been involved in cryptography and blockchain security since the early days of Bitcoin, cited the rapid advancement of artificial intelligence as the primary catalyst for their change of heart. “AI has made it trivial to find and exploit vulnerabilities that were previously hidden. Smart contract audits, even the most thorough ones, can no longer keep pace with what an AI can discover in minutes.”
The comment reflects a growing unease among security researchers. While DeFi has always carried risks—flash loan attacks, oracle manipulation, and reentrancy bugs have plagued the space—the infusion of generative AI and autonomous agents has dramatically escalated the threat landscape. Automated exploit bots can now scan codebases, model potential attack vectors, and execute complex multi-step exploits without human intervention.
The State of DeFi Security
Decentralized finance allows users to lend, borrow, trade, and earn yields using cryptocurrencies without intermediaries. The total value locked (TVL) in DeFi peaked at over $200 billion in late 2021 but has since dropped to around $80 billion as hacks and market downturns eroded confidence. According to data from blockchain analytics firm Chainalysis, DeFi platforms suffered $3.5 billion in losses from hacks in 2025 alone, a 40% increase from the previous year.
The pioneer’s blanket condemnation is particularly striking because they were once an ardent proponent of the technology. In a 2021 blog post, they wrote: “DeFi represents the most innovative use of blockchain technology since Bitcoin. It has the potential to democratize finance.” Now, they paint a much darker picture. “The very features that make DeFi attractive—open-source code, composability, programmable money—also make it a playground for AI-driven attacks. Every new protocol is an invitation to be exploited.”
Historical Context: The Rise of DeFi Attacks
The history of DeFi is littered with high-profile incidents. The 2016 DAO hack, which drained $60 million worth of Ether, was an early warning. But the pace accelerated in 2020 with the “DeFi Summer.” Since then, attacks have grown in sophistication. The 2022 Wormhole bridge hack ($320 million), the 2023 Euler Finance exploit ($197 million), and the 2024 Mango Markets incident ($114 million) are just a few examples. Many of these attacks leveraged smart contract vulnerabilities that could have been caught by more advanced tooling.
AI has turned a slow, manual process into an automated arms race. Researchers at MIT demonstrated in late 2025 that a GPT-class model could identify critical bugs in Solidity code with 95% accuracy, compared to 70% for the best human auditors. The same model could also generate exploit code ready for deployment. “We’ve gone from a world where you needed a team of expert engineers to break a protocol to one where a single person with an AI subscription can do it,” warned the pioneer.
Industry Reaction
The response from the DeFi community has been mixed. Some dismissed the warning as hyperbole, pointing out that the majority of protocols have never been hacked. Others, however, acknowledged the trend. “The security game has fundamentally changed,” said Sarah Chen, a leading smart contract auditor at Trail of Bits. “We are now battling AI with AI. It’s an arms race, and we are losing.” Chen noted that many audit firms are now investing in AI-based fuzzing and formal verification tools, but the attackers are always a step ahead.
Notable DeFi protocols like Uniswap and Aave have seen increased scrutiny. While no major exploits have occurred on these platforms in 2026, smaller protocols have been hit repeatedly. The pioneer pointed to the collapse of the lending platform Salarium in March, where an AI-driven flash loan attack drained $45 million in a matter of seconds. “That was the turning point for me. I realized that no amount of decentralized governance or emergency pause mechanisms can stop an exploit that happens in one block.”
Broader Implications
The warning has implications beyond DeFi. If AI can make decentralized finance fundamentally unsafe, it raises questions about the entire blockchain ecosystem. NFTs, gaming tokens, and even layer-1 chains could be next. Some experts argue that the solution lies in more sophisticated on-chain monitoring and real-time threat detection. Others advocate for a return to centralized oversight, which many in the crypto community view as antithetical to the movement’s ethos.
Regulators are also paying attention. The U.S. Securities and Exchange Commission (SEC) has reportedly opened an investigation into several DeFi exchanges following the pioneer’s comments. Senator Cynthia Lummis, a long-time crypto supporter, said in a statement: “If the technology cannot be secured, we must consider whether it has any place in our financial system.” Meanwhile, the European Union is fast-tracking its AI and blockchain regulation framework, with provisions that could require all DeFi applications to undergo mandatory AI-assisted security audits.
The pioneer’s final thought was somber: “I used to believe that code was law. But now, I believe that AI is the law—and it’s not on our side. Unless we find a way to make smart contracts inherently AI-resistant, DeFi as we know it is doomed.”
Source: Gizmodo News