The Artificial Intelligence Cyber Challenge (AIxCC) is a two-year cybersecurity competition led by DARPA in partnership with ARPA-H and other U.S. government agencies. The challenge brings together leading researchers and developers in artificial intelligence and cybersecurity to create AI-enabled systems capable of autonomously identifying and fixing vulnerabilities in critical open-source software.
Visit the official AIXCC website to learn more about the competition, explore detailed challenge information, and discover more CRS systems developed by other participants.
Visit AIXCC Website →Our project is open source and available on GitHub. You can find the source code, documentation, and contribute to the development of the CRS Leaderboard and related tools.
Explore comprehensive performance rankings of leading AI models evaluated on the AIXCC security challenge benchmarks. Track how different models perform in vulnerability detection and patch generation across various programming languages and challenge types.
View on GitHub →Access the complete collection of security challenges used in the AIXCC competition, including C and Java-based vulnerabilities.
View Challenges Repository →Discover our complete CRS implementation, and stay updated with the latest developments. To reproduce the results, please switch to the branch "qx".
View CRS Repository →zesheng@tamu.edu
qingxiao@tamu.edu
https://o2lab.github.io/