Researchers Used a Public OpenAI Model to Build a $250,000 Chrome Full-Chain Exploit in Six Days

Security researchers demonstrated that AI agents powered by large language models can drastically compress the timeline for developing sophisticated exploit chains, turning what once required weeks or months of expert effort into a six-day task costing roughly $250,000. By chaining three previously unknown (zero-day) Chrome vulnerabilities, the team achieved arbitrary code execution on Windows simply by loading a web page — a critical threat vector affecting virtually every enterprise environment. This development signals a fundamental shift in the threat landscape: defenders can no longer rely on the assumption that complex exploit development is a natural bottleneck for attackers. Organizations must treat zero-day vulnerability windows as significantly shorter than historical norms and accelerate both detection and patching capabilities accordingly.