Radware announced the launch of its LLM Firewall on November 18, 2025, adding a new layer of protection to its Cloud Application Protection Services. The firewall is the first phase of the company’s broader agentic AI protection strategy and is integrated into the EPIC‑AI framework, which infuses intelligence across Radware’s cloud‑security platform.
The LLM Firewall is model‑agnostic and delivers real‑time, AI‑based defense for both the prompt and the response of generative‑AI models. It blocks prompt injection, jailbreaks, and resource‑abuse attacks while ensuring compliance with data‑protection regulations such as GDPR and HIPAA. By protecting the entire request‑response cycle, the firewall reduces the attack surface that attackers can exploit in natural‑language interfaces.
Radware’s cloud‑ARR growth underpins the strategic importance of the new product. The company reported $89 million in cloud ARR in the most recent quarter, up 24% year‑over‑year, and is targeting $100 million by year‑end 2025. The LLM Firewall expands the company’s footprint into the high‑growth generative‑AI security market, a segment that is expected to drive double‑digit revenue growth for Radware as it leverages its existing cloud‑security customer base.
The generative‑AI security market is projected to grow rapidly, but the fact‑check report does not provide a specific market‑size estimate. Radware’s entry into this space is therefore positioned to capture a share of a market that is expanding faster than many traditional security segments.
Key competitors in the LLM‑security space include other cloud‑security vendors that offer AI‑based threat detection, but Radware differentiates itself by integrating the firewall directly into its existing Cloud Application Protection Services. This integration allows customers to add LLM protection without disrupting current workflows or requiring new infrastructure.
No early‑adoption data or pilot‑program results were disclosed in the fact‑check report. Radware has indicated that it will roll out pilot programs to gather customer feedback and refine the product, but specific outcomes are not yet available.
The agentic AI protection strategy is described as a multi‑phase initiative. The LLM Firewall represents the first phase, with future phases expected to include AI model monitoring, policy enforcement, and advanced threat analytics, although the company has not yet released detailed plans for these subsequent components.
Constance Stack, Radware’s Chief Growth Officer, emphasized the importance of enforcing security at the prompt level: “AI security must be enforced at the prompt in order to defend against prompt injection, jailbreaks, and resource abuse.” CEO Roy Zisapel added that the company is “leveraging AI‑powered innovations to enhance our platform and competitive standing.”
The launch of the LLM Firewall positions Radware to capture a growing segment of the AI‑security market while reinforcing its cloud‑ARR growth strategy. By offering a model‑agnostic, real‑time defense that aligns with regulatory requirements, Radware is poised to meet the evolving security needs of enterprises adopting generative AI.
The content on BeyondSPX is for informational purposes only and should not be construed as financial or investment advice. We are not financial advisors. Consult with a qualified professional before making any investment decisions. Any actions you take based on information from this site are solely at your own risk.