F5 introduced enhanced threat intelligence resources enabling enterprises to measure and compare risk profiles across popular AI model. The F5 Labs team launched CASI and ARS leaderboards with standardized, monthly benchmarks backed by research. These benchmarks combine real-time attack intelligence with expert analysis of evolving AI attack vectors.
Expanded AI Vulnerability Intelligence
Following its acquisition of CalypsoAI, F5 expanded one of the industry’s largest AI vulnerability libraries. The library adds over 10,000 new attack prompts monthly using more than a year of data. Enterprises can evaluate and select AI model based on real-world application security threats before production deployment.
“Deploying unverified AI models into critical infrastructure is not innovation; it is negligence. Organizations need a way to continuously quantify resilience. F5 Labs AI Leaderboards offer that standard. These rankings isolate specific weaknesses in the model layer, giving security teams the intelligence they need to govern inference and block attacks before they happen,” said Kunal Anand, Chief Product Officer at F5.
Leaderboards enhance visibility and help teams identify vulnerabilities across environments. They strengthen defenses alongside the F5 Application Delivery and Security Platform. The platform safeguards APIs, protects data, prevents DDoS attacks, and accelerates DevSecOps automation.
Advancing Model Validation and AI Governance
Rapid AI integration requires robust security validation of enterprise models. F5 Labs established metrics measuring the least resistance paths and minimum compute resources for attacks. CASI evaluates average performance, risk-to-performance ratio, and cost of security. Meanwhile, ARS measures resilience against sustained, adaptive, multi-step AI agent attacks. F5 AI Guardrails and F5 AI Red Team complement these leaderboards. They apply governance controls and simulate adaptive security testing conditions.
“F5’s existing traffic inspection and logging capabilities are expanded with the acquisition of CalypsoAI, adding runtime AI governance capabilities and visibility into AI system behavior. Paired with their security research arm, F5 Labs, that updates CASI and ARS rankings every month, F5 customers are provided with both a useful set of security metrics to consider when selecting an AI model, as well as the tools needed to constrain how AI systems are used, what data they may access, and how outputs are generated,” said Grace Trinidad, Research Director, AI Security & Trust, IDC.
To explore how Security Operations Centers (SOC) play a crucial role in defending against modern cyber threats, read our latest SOC News.
Source: Businesswire
