Vitalik: AI Won't Wreck Crypto Security—It Might Even Math Its Way to Proofs 🧠

Ethereum co-founder Vitalik Buterin pushed back against the growing belief that artificial intelligence will hand hackers an unbeatable edge, arguing instead that advanced AI could help make crypto software provably secure. In a Wednesday post on X, Buterin wrote that crypto holders—including himself, with roughly 90% of his net worth in crypto—are betting that defensive tooling can keep pace with offensive advances. "It's an increasingly common take that AI hacking means cybersecurity is doomed," Buterin wrote. "I disagree. I think cybersecurity is naturally defense-favoring once people get their shit together."

Buterin said AI could make formal verification—the practice of using mathematical proofs to confirm software behaves as intended—practical even for complex systems. "If AI can prove Navier-Stokes and FLT, then AI can prove the statement 'this program is secure' as a mathematical theorem," he wrote. "Even if the program is very complicated." He cautioned that the hard part is not the verification itself but defining what "secure" actually means before any proof can be written.

His comments follow a string of incidents in which AI has already altered the offense-defense balance across the crypto sector. In May, security researcher Taylor Hornby used Anthropic's Claude Opus 4.8 to uncover a four-year-old flaw in Zcash's Orchard privacy pool that could have enabled unlimited, undetectable counterfeiting of ZEC; developers said they found no evidence of exploitation before patching it in June. The following month, attackers drained Coldcard wallets through a years-old firmware flaw that weakened seed generation, ultimately stealing roughly $130 million in $BTC, with device maker Coinkite saying AI likely helped identify the bug. Ethereum Foundation researchers also said in July that AI agents had surfaced vulnerabilities in critical network infrastructure.

By August, the impact had spread across Bitcoin software. Boltz suspended its swap service, saying suspected attackers were finding flaws faster than developers could patch them, while Core Lightning confirmed that several vulnerabilities flagged in AI-generated reports were genuine. In response, the volunteer Bitcoin Red Team used AI-assisted audits to flag 4,962 potential vulnerabilities across 390 projects. Buterin said AI can now verify entire programs rather than isolated components, an approach he said Ethereum plans to pursue over the next several years, writing that "there is no future for blockchains—especially blo