en.Wedoany.com Reported - F5 recently announced the expansion of its artificial intelligence security strategy by acquiring SurePath AI and launching a new AI Security Platform, addressing the reality that enterprise AI adoption has outpaced security governance. The acquisition provides the company with network-based discovery capabilities aimed at uncovering both authorized and unauthorized AI usage. In many enterprises, shadow AI has evolved into a structural challenge, described by security leaders as a significant unknown factor in the environment.

AI systems currently operate within regulated networks, behind APIs, and in agents that require no human supervision. F5's Chief Product Officer noted in the announcement that most current AI security tools are merely shells for chatbots, and enterprises need controls that match the reality of their actual infrastructure. According to McKinsey 2024 data, 72% of organizations used AI in at least one business line that year, up from 55% the previous year. This rapid expansion explains why runtime oversight is becoming a core component of enterprise AI strategies. Meanwhile, the U.S. National Institute of Standards and Technology (NIST) 2023 AI Risk Management Framework and the 2024 EU AI Act introduced structured risk management approaches and compliance obligations varying by system risk level, respectively. Analysts from Gartner and the European Union Agency for Cybersecurity (ENISA) have listed prompt injection and data leakage as top threat categories.

F5 positions the new platform as a continuous loop to align with regulatory pressures. The architecture integrates governance, discovery, testing, and runtime protection modules, connected through an observability layer. The governance module translates risk tolerance and regulatory requirements into enforceable boundaries covering prompts, outputs, tool usage, and data access. The discovery component, enhanced by SurePath AI, aims to capture AI activities without inserting new code into applications and classify usage by intent and use case. The testing module stress-tests systems against over 140,000 attack patterns. In terms of runtime protection, the organization reports that its guardrails achieved a security efficacy of 98.2% in independent testing.

Considering varying enterprise deployment preferences, F5 emphasizes that the AI Security Platform can be deployed on-premises, in air-gapped environments, private clouds, or across hybrid and public cloud settings to address practical issues such as data residency requirements and sovereignty rules. AI observability differs from traditional logging through advanced traceability, which is particularly important for AI agents that autonomously authenticate, invoke tools, and access data. The organization's 2026 State of Application Strategy Report indicates that 98% of surveyed enterprises are preparing for agentic AI. As agent systems can autonomously link actions, continuous monitoring has become a necessity beyond static compliance checks.

This platform launch comes amid a shifting competitive landscape in AI security. Vendors such as Palo Alto Networks and Microsoft have also integrated AI security controls into their cloud and application security stacks. Leveraging its expertise in application delivery and security, F5 strengthens its focus on discovery through the acquisition of SurePath AI to address the ongoing expansion of shadow AI. This release reflects a broader industry consensus: AI has become a core component of application architecture, and organizations need controls that match the speed of model operations to prevent unauthorized model usage or unexpected decisions by autonomous agents. For enterprise leaders seeking to manage risk in an AI-pervasive network environment, this continuous monitoring approach integrating discovery, governance, testing, and protection offers a pragmatic path forward.

This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com