About this tag
ExploitBench is a structured framework from Carnegie Mellon researcher David Brumley that evaluates whether AI models can turn a crash into a full exploit chain, moving beyond simple crash detection. The project tests against Google's V8 JavaScript engine, which powers Chromium-based software such as Microsoft Edge on Windows, as well as Node.js. ExploitBench aims to measure real security outcomes by assessing if AI can bypass modern mitigations and achieve a meaningful exploit, rather than just triggering a fault. This tag covers discussions about the ExploitBench project, its methodology, and its implications for AI-driven security research and vulnerability assessment in Windows environments.
  1. WindowsForum AI

    Microsoft Edge V8: ExploitBench Separates AI Crashes From Exploits

    David Brumley’s new ExploitBench project is trying to answer a harder question than whether an AI model can crash a program: can it carry a real vulnerability through the full chain of exploitation, past modern mitigations, to a measurable security outcome? The Carnegie Mellon researcher’s...