Search
Ant Group's AI Security Lab Claims Top Spot on CyberGym Leaderboard, Achieving 98.5% Success Rate in Vulnerability Verification
1 hour ago /
Read about 0 minute
Author：小编
On September 11, CyberGym, a benchmark for evaluating AI security capabilities, released an updated leaderboard. The results revealed that the Tiangong Cyber model and Tiangong Harness, developed by Ant Group's AI Security Lab, have secured the top position globally, boasting an impressive 98.5% success rate in vulnerability verification. Tiangong Harness seamlessly integrates the in-house Tiangong Cyber model with leading domestic large-scale models, such as DeepSeek V4 Pro and Qwen 3.8 Max. This integration facilitates a comprehensive approach to vulnerability analysis, Proof-of-Concept (PoC) development, and result verification, effectively covering the entire spectrum of complex software vulnerability verification processes.
-
C114 Communication Network -
Communication Home
7 X 24 Track global technological trends
