A top crypto security executive warned this week that artificial intelligence has become 'superhuman' at hacking, calling decentralized finance 'unsafe' as losses pile up. The executive — whose identity is known inside the industry but not disclosed in the warning — said AI coding agents are now able to find and exploit smart contract flaws faster than human auditors can patch them. The warning lands as DeFi's total value locked continues to fall, with hack after hack chipping away at user confidence.
What the executive said
In a stark assessment shared privately with industry peers, the executive described AI's offensive capabilities as a 'quantum leap' that defenders haven't caught up with. 'We're seeing AI agents that can reverse-engineer a contract, find every logical hole, and craft an exploit in minutes,' the executive reportedly said. The key claim: smart contracts are now 'fatally vulnerable' because AI can model attack vectors humans would miss. The executive did not name specific protocols, but the framing was broad — no DeFi project is safe, they argued.
TVL keeps sliding
DeFi's total value locked has been dropping for weeks, and the timing of this warning doesn't help. Hacks have become almost routine: another exploit hit a lending protocol earlier this month, and a cross-chain bridge was drained last week. Each incident drives more liquidity out. The executive's message reinforces what many users already suspect — that the security model of DeFi is broken, and AI is making it worse.
Why AI changes the game
Traditional smart contract audits rely on humans checking code line by line. AI agents don't get tired, don't overlook edge cases, and can test millions of permutations in seconds. The executive said the gap between attack speed and defense speed is now 'unbridgeable' without a fundamental rethink. They didn't offer a solution — just a warning that the industry needs to treat AI as an adversary that never sleeps.
What comes next
No regulator has responded publicly. No major DeFi platform has announced new AI-driven defenses. The executive's warning is a shot across the bow, but it's unclear who will act on it. For now, the industry is left with a hard question: if AI can break any contract, what's left to build on?
