Check Point® Software Technologies Ltd. (NASDAQ: CHKP), a leader in cyber security solutions, announced the launch of its AI Factory Security Architecture Blueprint, designed to enhance the security of private AI infrastructure. This comprehensive reference architecture aims to protect against vulnerabilities from the hardware to application layers, responding to the significant investments enterprises and neo-clouds are making in private AI environments.
“AI infrastructure has become one of the most valuable and vulnerable assets in the enterprise,” said Nataly Kremer, Chief Product Officer at Check Point. She emphasized that the AI Factory Security Blueprint is intended to safeguard these investments from the ground up, ensuring security is integrated into the architecture rather than treated as an afterthought.
The AI data centre, central to enterprise operations, has evolved into a critical yet exposed asset as organizations increasingly build private AI environments to protect intellectual property and reduce reliance on public cloud services. These private setups are accumulating high-performance GPU clusters, training pipelines, and proprietary models, creating substantial financial commitments that outpace the development of corresponding security measures.
Unlike traditional data centres, AI computing environments incorporate GPU clusters, distributed training pipelines, and real-time inference APIs, generating unique attack surfaces that standard security tools struggle to address. Threats include data poisoning, model theft, and breaches through lateral movements within Kubernetes clusters. As organizations ramp up their AI capabilities, the urgency for robust security measures becomes paramount.
The Check Point AI Factory Security Blueprint proposes layered protection at four distinct levels. At the perimeter layer, the Check Point Maestro Hyperscale Firewall offers Zero Trust Network Access (ZTNA) and scalable policy enforcement, effectively managing traffic between external users and enterprise networks. The application and large language model (LLM) layer features Check Point AI Agent Security, which protects inference APIs and LLM endpoints from sophisticated attacks that traditional web application firewalls cannot effectively mitigate.
Furthermore, the AI infrastructure layer benefits from a collaboration with NVIDIA, embedding Check Point’s firewall and threat prevention capabilities directly into NVIDIA BlueField data processing units (DPUs). This integration allows for high-performance security that does not tax CPU or GPU resources, enabling real-time traffic inspection and threat detection without compromising performance. The workload and container layer employs micro-segmentation solutions to control traffic within Kubernetes clusters, effectively isolating any compromised containers to prevent wider disruptions.
This security framework aligns with the Cybersecurity and Infrastructure Security Agency’s (CISA) principle that AI must be Secure by Design, advocating for security measures embedded from inception across all layers of architecture. By enforcing Zero Trust at every interaction, Check Point ensures that user access and API calls are continuously validated and authenticated.
Moreover, the blueprint aligns with AI governance frameworks such as the NIST AI Risk Management Framework and Gartner AI TRiSM, providing essential traceability and auditability to help organizations comply with emerging regulations, including the EU AI Act and GDPR.
As the demand for AI capabilities surges, the necessity for robust security measures in private AI infrastructures will only grow. Check Point’s newly introduced AI Factory Security Blueprint offers a proactive approach to safeguarding these critical assets, underscoring the importance of embedding security into the architecture from the outset.
See also
Alibaba Launches XuanTie C950 Chip, Tripling AI Performance with RISC-V Architecture
Tiiny AI Pocket Lab Launches: 80GB RAM Device Enables Local 120B Model Inference
Prince George’s County Invests $1M in AI and Quantum Computing Partnership with UMD
UAE’s Strategic Purchase of US GPUs Positions American Tech as Global Standard, Says Trump’s AI Adviser
AI Search Engineers Launches Answer Engine Optimization to Enhance Brand Visibility in AI Systems
