Radware Ltd. (RDWR) has launched a new LLM Firewall designed to protect enterprise generative AI systems from data leakage, prompt injection, and unauthorized access. The solution targets growing cybersecurity risks in AI-driven environments.
- Radware Ltd. (RDWR) launched the LLM Firewall to secure generative AI systems
- Supports models including Llama 3 and GPT-4 with real-time threat detection
- Achieves 98.4% detection accuracy in internal testing
- Handles up to 12,000 queries per second
- Targets compliance with HIPAA, GDPR, and PCI-DSS
- Emerging market for AI security projected to exceed $12B by 2027
Radware Ltd. (RDWR) has introduced the LLM Firewall, a specialized security product aimed at mitigating risks associated with large language model (LLM) deployments in enterprise environments. The solution is designed to monitor, filter, and enforce policies across AI-generated content in real time, addressing vulnerabilities such as prompt injection attacks, data exfiltration, and model misuse. The product integrates directly with existing AI infrastructure, offering automated threat detection and policy enforcement at scale. The LLM Firewall supports multiple LLM architectures, including open-source models like Llama 3 and proprietary platforms such as GPT-4. It leverages behavioral analysis and natural language understanding to identify malicious inputs and sensitive data patterns, with detection accuracy reported at 98.4% in internal benchmark tests. The system processes up to 12,000 queries per second, making it suitable for high-volume enterprise applications. Early adopters include global financial institutions and healthcare providers, where data compliance and regulatory requirements are stringent. These organizations are using the firewall to meet standards such as HIPAA, GDPR, and PCI-DSS, ensuring that AI interactions do not compromise sensitive customer or operational data. The launch comes amid rising concerns over AI security, with a 2025 industry survey showing that 68% of enterprises experienced at least one AI-related security incident in the past year. Radware's new solution positions the company as a key player in the emerging AI security market, with projected revenue from AI-specific cybersecurity tools expected to exceed $12 billion by 2027.