Check Point Launches AI Factory Security Blueprint to Safeguard Enterprise AI

The Check Point Software Technologies  has unveiled a new security framework called the AI Factory Security Architecture Blueprint, designed to protect private artificial intelligence infrastructure across all layers — from hardware and data pipelines to applications and large language model (LLM) environments.

The announcement highlights the growing security risks facing modern AI data centers, which increasingly rely on high-performance GPU clusters, distributed training systems, and real-time inference APIs. According to the company, these environments have become highly valuable yet vulnerable assets, with emerging threats including data poisoning, model theft, prompt injection, and supply chain attacks.

To address these risks, the blueprint introduces a layered “security-by-design” approach that integrates protection directly into AI infrastructure. It combines Check Point’s cybersecurity technologies with NVIDIA’s BlueField data processing capabilities to deliver embedded, high-performance security across enterprise AI systems.

At the perimeter level, the architecture uses Check Point Maestro Hyperscale Firewall to enforce Zero Trust Network Access, segmentation, and scalable policy enforcement. At the application layer, Check Point AI Agent Security protects LLM endpoints and APIs against prompt injection, data leakage, and adversarial attacks, extending safeguards beyond traditional web application firewalls.

At the infrastructure level, the solution integrates directly with NVIDIA BlueField data processing units using the NVIDIA DOCA platform, enabling hardware-accelerated security, real-time threat detection, and traffic inspection without burdening CPU or GPU resources. At the workload layer, integration with Kubernetes micro-segmentation tools helps prevent lateral movement within clusters and isolates compromised containers.

The framework aligns with global security and governance standards such as the CISA Secure-by-Design principles, the NIST AI Risk Management Framework, and Gartner AI TRiSM. It is also designed to support compliance with major regulatory requirements including the EU AI Act, GDPR, HIPAA, PCI-DSS, and ISO 42001.

Nataly Kremer, Chief Product Officer at Check Point, said AI infrastructure has rapidly become one of the most valuable and exposed enterprise assets, stressing that security must be embedded from the ground up rather than added after deployment.

Overall, the blueprint positions Check Point’s approach as a comprehensive model for securing next-generation AI factories, ensuring enterprises can safely scale AI innovation while maintaining resilience, compliance, and operational trust.