F5 introduces AI remediate for model vulnerability protection

F5 has announced major updates to its Application Delivery and Security Platform, introducing new capabilities for protecting AI-driven applications, expanding zero trust access controls, and preparing for post-quantum cryptography, reports IT Brief Australia.

The enhancements span BIG-IP, NGINX, and Distributed Cloud Services, including F5 AI Remediate, which automates the creation of runtime guardrails from identified AI model vulnerabilities. Kunal Anand, Chief Product Officer, stated that "security teams do not need more alarms. They need fewer gaps," emphasizing that ADSP "closes the loop from finding risk to enforcing protection."

New Distributed Cloud Bot Defence capabilities enable clear separation of humans, bots, and AI agents, allowing only trusted verifiable agents to interact with applications. BIG-IP Access Policy Manager evolves into BIG-IP Zero Trust Access, supporting Identity Aware Proxy and post-quantum cryptography-ready TLS and VPN tunnelling. F5 Insight for ADSP provides end-to-end observability using OpenTelemetry, turning scattered telemetry into actionable intelligence. API security enhancements include out-of-band discovery across multiple data planes and deployable options for air-gapped environments. The updates address agentic AI workloads, Model Context Protocol traffic, and HTTP/3 attack patterns.