F5 has launched a new AI Security Platform aimed at giving enterprises greater control over the risks associated with rapidly expanding AI deployments, alongside the acquisition of AI visibility specialist SurePath AI.
F5 says the platform is designed to provide continuous visibility, governance and protection across AI applications, models, agents and the APIs that connect them. The move reflects growing concern among enterprises over “shadow AI” and the increasing autonomy of AI systems operating within corporate environments.
F5 said the platform extends its existing Application Delivery and Security Platform (ADSP) strategy to AI workloads, supporting deployment across on-premises, air-gapped, private cloud, hybrid and public cloud environments. This flexibility is intended to address data sovereignty and regulatory requirements, particularly in highly regulated sectors.
The launch comes as organisations grapple with the operational and security implications of AI adoption. According to F5’s 2026 State of Application Strategy report, 88% of organisations have experienced at least one AI-related operational or security challenge. Risks include prompt injection, data leakage and AI agents acting beyond authorised parameters.
“Most AI security today is a wrapper around a chatbot. That is not security,” said Kunal Anand, chief product officer at F5. “Enterprises run AI inside regulated networks, behind APIs, and across agents that authenticate and act on their own.”
A key component of the new platform is the integration of SurePath AI, which F5 has acquired to enhance its capabilities in AI discovery and risk detection. SurePath’s network-based technology enables organisations to identify AI usage across their environments, including unsanctioned tools, without requiring direct integration at the application level. It also classifies the intent behind AI workflows and tracks agent activity, providing a real-time visibility layer for security teams.
The platform is structured around four core pillars—governance, discovery, testing and runtime protection—supported by an overarching observability layer. Together, these create a continuous security lifecycle rather than a one-off compliance process.
Key features include policy-based governance controls, continuous discovery of AI applications and agents, and pre-deployment testing against a large database of attack patterns. Runtime protection capabilities allow organisations to enforce guardrails at the point of interaction, with F5 citing efficacy rates of over 98% in blocking threats such as prompt injection and data exfiltration.
The platform also delivers end-to-end auditability, enabling organisations to track AI interactions and maintain compliance with regulatory standards.
F5 said the offering is designed to address the growing complexity of agentic AI, as autonomous systems increasingly authenticate, access data and execute actions independently. With 98% of organisations preparing for agentic AI, the company warned that adoption is outpacing the controls needed to manage associated risks.
