Check Point Unveils AI Defense Plane with 50ms Runtime Control and 4-Layer AI Factory Blueprint
Check Point launched the AI Defense Plane, a unified control plane delivering sub-50ms runtime security across employee AI use, applications and autonomous agents in over 100 languages. It also unveiled the AI Factory Security Blueprint, a four-layer architecture securing private AI data centers from GPU clusters and inference endpoints.
1. Launch of AI Defense Plane
Check Point introduced the AI Defense Plane to govern and secure agentic AI operations across enterprises. Built on ThreatCloud AI and enhanced by recent Lakera and Cyata acquisitions, its AI-native security engine analyzes millions of interactions, delivering adaptive protection in under 50 milliseconds across more than 100 languages. The platform comprises three modules—Workforce AI Security, AI Application & Agent Security, and AI Red Teaming—to provide discovery, governance, runtime control and continuous adversarial testing for employee AI use, embedded agentic systems and production workflows.
2. Release of AI Factory Security Blueprint
Check Point published the AI Factory Security Blueprint, a vendor-tested reference architecture for private AI infrastructure. The blueprint defines four protection layers—perimeter using Maestro Hyperscale Firewall, application and LLM APIs defense, hardware-accelerated inspection via NVIDIA BlueField DPUs, and container microsegmentation—to secure GPU clusters, training pipelines and inference services while aligning with NIST, EU AI Act and other regulatory frameworks.