F5 tackles AI security with new platform extensions (www.networkworld.com)

F5 has announced significant enhancements to its Application Delivery and Security Platform, specifically targeting AI security and multi-cloud operations. On January 14, the company unveiled F5 AI Guardrails and F5 AI Red Team, tools developed after its acquisition of CalypsoAI in 2025. These products are designed to protect AI models from a unique set of security threats that traditional network defenses cannot address. F5 AI Guardrails functions as a proxy, intercepting and examining user prompts and model outputs to block harmful interactions and ensure compliance with regulations like GDPR. Meanwhile, F5 AI Red Team automates adversarial testing, continuously updating a database of vulnerabilities to enhance the Guardrails' policy enforcement. The significance of these announcements lies in the evolving nature of AI security, which demands different strategies compared to traditional cybersecurity measures. As AI models become increasingly integrated into enterprise applications, they present novel attack surfaces that require solutions to defend against issues like prompt injection and data leakage. F5’s dual offerings not only bolster the security infrastructure for AI applications but also illustrate a shift toward a more comprehensive and adaptive security paradigm in the AI/ML community, paving the way for secure and scalable AI operations.