F5实验室每月发布AI安全头板,采用新的衡量标准来评估安全模式和对先进攻击的抵抗能力。 F5 Labs launches monthly AI security leaderboards using new metrics to assess model safety and resistance to advanced attacks.

F5实验室利用AI安全指数(CASI)和抗药性计分(ARS)综合指数(ARS)每月推出AI安全领导板,以帮助各组织评估AI安全模式和复原力。 F5 Labs has launched monthly AI security leaderboards using the Comprehensive AI Security Index (CASI) and Agentic Resistance Score (ARS) to help organizations assess AI model safety and resilience. 这些工具评估关于性能、安全成本和抵御适应性、目标驱动的AI攻击的模型。 Updated with over 10,000 new attack prompts monthly and backed by a year of threat data, the tools evaluate models on performance, security cost, and resistance to adaptive, goal-driven AI attacks. 在F5收购CalypsoAI后,这项举措为安全部署AI提供了独立基准,补充了运行时治理和威胁测试的现有工具。 Built on F5’s acquisition of CalypsoAI, the initiative provides independent benchmarks to guide secure AI deployment, complementing existing tools for runtime governance and threat testing.