Check Point Software Technologies Ltd. (NASDAQ: CHKP), a global cybersecurity solutions, announced a collaboration with NVIDIA to deliver an integrated security solution for AI factories. The company revealed that AI Cloud Protect is now available for on-premises enterprise deployments, securing AI model development, agentic AI applications, and inference workloads without affecting performance.

The solution has been validated on NVIDIA RTX PRO Servers, enabling enterprises to deploy AI securely, from data center to cloud, using NVIDIA BlueField technology. Nataly Kremer, Chief Product Officer at Check Point, said, “With NVIDIA, we’re making AI factories secure by design, protecting models, data, and infrastructure without slowing innovation.”

Recent studies show that AI systems face unprecedented security risks. Check Point data revealed that 1 in every 80 GenAI prompts exposes sensitive data. Meanwhile, a Gartner report found that 32% of organizations experienced AI attacks involving prompt manipulation, while 29% faced attacks on GenAI infrastructure in the past year.

AI Cloud Protect runs on NVIDIA BlueField-3 DPUs and delivers full-stack security without consuming CPU or GPU resources. It uses network-level protection to prevent unauthorized access, reduce data poisoning, and stop model exfiltration. At the host level, the solution leverages NVIDIA DOCA Argus to monitor all running processes, helping detect and prevent malicious activities in AI nodes and publicly downloaded Large Language Models (LLMs).

David Reber, Chief Security Officer at NVIDIA, added, “Security is essential for the next generation of AI infrastructure. Our collaboration with Check Point integrates BlueField acceleration and the NVIDIA DOCA Argus framework into AI Cloud Protect, enabling enterprises to deploy AI confidently.”

The platform will also run on the upcoming NVIDIA BlueField-4 DPU, which promises up to six times more compute power and double the network throughput for AI inference and storage, supporting scalable and secure AI operations at high speed.

Check Point also offers a comprehensive security portfolio, including CloudGuard Web Application Firewall with Lakera and Infinity GenAI Protect, which safeguard enterprise AI applications against prompt injection, jailbreaking, and sensitive data leaks. Early pilots of AI Cloud Protect are underway with financial services customers and partners such as World Wide Technology (WWT).

Chris Konrad, VP of Global Cyber at WWT, stated, “The combination of Check Point’s AI Cloud Protect and NVIDIA BlueField acceleration delivers enterprise-grade protection for AI workloads, from model training to inference, without compromising performance.”

Key AI security challenges include:

• Expansive attack surfaces from massive compute and data pipelines
• Threats like prompt injection, jailbreaking, and model poisoning
• Risks from ungoverned employee AI use

AI Cloud Protect addresses these with:

• Protection across network, host, and application layers
• Zero performance impact via NVIDIA BlueField-accelerated security
• Unified management and automation for thousands of AI nodes

Related post

View all