Check Point Unveils AI Factory Security Blueprint to Strengthen Protection of AI Infrastructure from GPU Servers to LLM Prompts
DUBAI, UAE – Check Point® Software Technologies Ltd. (NASDAQ: CHKP), a recognized leader in cybersecurity solutions, has introduced the AI Factory Security Architecture Blueprint. This comprehensive framework serves as a vendor-tested reference architecture aimed at securing private AI infrastructures from the hardware layer to the application layer. By leveraging Check Point’s advanced firewall and AI security technologies, alongside NVIDIA’s BlueField data processing capabilities, the blueprint ensures security is integrated at every level of the AI factory and data center.
Nataly Kremer, Chief Product Officer at Check Point, emphasized the significance of AI infrastructure, stating that it has become one of the most valuable yet vulnerable assets within enterprises. The AI Factory Security Blueprint is designed to protect these investments from the ground up, ensuring that security is not an afterthought but a foundational element throughout the entire stack.
The Strategic Importance of AI Data Centers
AI data centers have emerged as critical components of enterprise infrastructure, characterized by their strategic value and vulnerability. Organizations are increasingly building private AI environments to safeguard intellectual property, comply with sovereignty requirements, and mitigate public cloud costs. This rapid development has led to the accumulation of GPU clusters, training pipelines, inference workloads, and proprietary models, all of which represent significant investments. However, the pace of security architecture development has struggled to keep up with these advancements.
Unlike traditional data centers, AI computing environments integrate high-performance GPU clusters, distributed training pipelines, large-scale data lakes, and real-time inference APIs. This complexity creates attack surfaces that conventional security tools are ill-equipped to address. Threats include training data poisoning, model theft, lateral movement within Kubernetes namespaces, prompt injection targeting inference APIs, and supply chain vulnerabilities stemming from open-source dependencies.
Layered Protection Across Four Levels
The Check Point AI Factory Security Blueprint provides a multi-layered protection strategy across four distinct levels:
-
Perimeter Layer: The Check Point Maestro Hyperscale Firewall offers Zero Trust Network Access (ZTNA), virtual security group segmentation, and scalable policy enforcement at the entry point to the AI fabric. This layer effectively manages north-south traffic from external users, internet sites, and enterprise networks.
-
Application and LLM Layer: Check Point AI Agent Security protects inference APIs and LLM endpoints from threats such as prompt injection, data exfiltration, adversarial queries, and API abuse. This protection goes beyond traditional web application firewalls, with Check Point AI Agent Security integrated into Check Point Firewalls across various deployment models, including cloud, virtual, and appliance forms.
-
AI Infrastructure Layer: In collaboration with NVIDIA, Check Point has embedded its firewall and threat prevention capabilities directly into NVIDIA BlueField data processing units (DPUs) via the NVIDIA DOCA software platform. This integration delivers hardware-accelerated, inline security at the infrastructure level, ensuring high-performance AI prompt defense and traffic inspection without consuming CPU or GPU resources.
-
Workload and Container Layer: Check Point’s partnerships with third-party microsegmentation solutions facilitate micro-segmentation and east-west traffic control within Kubernetes clusters. This approach prevents lateral movement between inference namespaces and isolates compromised containers before they can spread.
Aligning with Security Principles and Regulatory Frameworks
The AI Factory Security Blueprint aligns with the Cybersecurity and Infrastructure Security Agency (CISA)’s principle that AI must be Secure by Design. This principle advocates for security to be embedded from the outset—in the fabric, hardware, and orchestration layer—rather than added as an afterthought to existing systems. Check Point’s architecture enforces a Zero Trust model at every interaction, ensuring that every user, API call, and service request is authenticated, authorized, and continuously validated.
Moreover, the blueprint is designed to comply with AI governance frameworks, including the NIST AI Risk Management Framework and Gartner AI TRiSM. It provides the necessary traceability, auditability, and policy enforcement to meet emerging regulations such as the EU AI Act, GDPR, HIPAA, PCI-DSS, and ISO 42001.
For further insights, refer to the original reporting on this development at Zawya.
About Check Point Software Technologies Ltd.
Check Point Software Technologies Ltd. is a global leader in cybersecurity, protecting over 100,000 organizations worldwide. The company is committed to securing enterprises’ AI transformation through a prevention-first approach and an open ecosystem architecture. Check Point enables organizations to block advanced threats, prioritize vulnerabilities, and automate security operations across complex digital environments. Its unified architecture simplifies protection across hybrid networks, multi-cloud environments, digital workspaces, and AI systems. Structured around four strategic pillars—Hybrid Mesh Network Security, Workspace Security, Exposure Management, and AI Security—Check Point delivers consistent protection and visibility across multivendor environments, allowing organizations to reduce risk, enhance efficiency, and accelerate innovation without increasing complexity.
For the latest cybersecurity developments, threat intelligence and breaking updates from across the Middle East: Middle East
