DARPA AI Cyber Challenge Announcement
DARPA Program Manager Perri Adams and industry leaders announce the AI Cyber Challenge (AIxCC) to automate software vulnerability discovery and remediation.
Summary
DARPA program manager Perri Adams is joined by industry and open-source leaders on the DEF CON 31 Main Stage to formally announce the launch of the AI Cyber Challenge (AIxCC).
[00:03] Adams introduces the initiative, which pairs DARPA with leading AI research laboratories—including Anthropic, Google, Google DeepMind, and OpenAI—alongside the Open Source Security Foundation (OpenSSF). The two-year competition is designed to develop novel Cyber Reasoning Systems (CRSs) that integrate advanced artificial intelligence and large language models with automated program analysis [09:20] to secure critical open-source software infrastructure.
[13:22] The panelists discuss the technical vision of AIxCC: moving beyond manual vulnerability research to create autonomous systems that can discover complex software vulnerabilities, verify exploitability, and automatically generate and test verified source code patches. Adams outlines the competition structure and substantial prize pool [17:27], featuring over $20 million in funding culminating in finals at DEF CON 33, inviting global security researchers and AI practitioners to participate.
(AI summary — Gemini Flash)
Related talks
Village Interviews·1 min
Andrew Carney
Andrew Carney from DARPA details the finals of the AI Cyber Challenge, where autonomous AI agents detect and repair software vulnerabilities.
Village Interviews·1 min
A preview of the multi-million-dollar finals of the DARPA AI Cyber Challenge (AIxCC), evaluating autonomous cyber reasoning systems.
Main Stage·42 min
DARPA AIxCC Finalists
Finalists from DARPA's AI Cyber Challenge present autonomous cyber reasoning systems designed to find, patch, and remediate critical software vulnerabilities.