Yet, despite such optimism, something unreal happened soon after OpenAI launched EVMbench. An exploit involving Claude Opus 4.6 raised serious concerns about the risks of “vibe-coded” smart contracts.

Robust security is critical in open networks where software controls user funds. Security strategies must evolve as AI’s role in smart contract evaluation grows.

Crypto protocols lost hundreds of millions of dollars to smart contract failures in 2025. Now, the OWASP Smart Contract ...

OpenAI and Paradigm have released EVMbench—a framework for evaluating AI agents' ability to find vulnerabilities in Ethereum smart contracts.

OpenAI, along with Paradigm and Ottersec, has released the EVMbench research paper, looking at how well different AI models ...

EVMbench is OpenAI’s attempt to see whether modern AI systems are up to the task of helping prevent smart contract issues.

OpenAI launches EVMbench with Paradigm to test AI on smart contract vulnerabilities and commits $10M to cybersecurity ...

Ethereum is unique among cryptocurrencies in that it is as much a commodity to be converted from ETH to USD as it is a global software platform that operates on blockchain technology. The ecosystem ...

While exploitation of security flaws in many smart contracts have become a perennial target of cyberattackers, more security firms are shining a spotlight on scams that use fraudulent or obfuscated ...

A critical smart-contract flaw in Balancer V2 allowed unauthorized withdrawals totaling over $100 million. The incident spanned multiple networks and assets, exposing systemic vulnerabilities in DeFi ...

Smart contracts have long been the backbone of blockchain’s promise. These self-executing programs run exactly as coded, without downtime, fraud, or third-party interference—at least in theory. But in ...

A $26 million exploit of the offline computation protocol Truebit stemmed from a smart-contract flaw that allowed an attacker to mint tokens at near-zero cost, highlighting persistent security risks ...