Check Point launches AI Defense Plane for Enterprise AI Security for organizations managing AI systems across workforce tools, applications, and autonomous agents.
The platform is designed as a unified control plane for managing how AI is connected, deployed, and operated across your business. This covers the risks that arise when AI systems move beyond content generation to tasks that involve accessing data, invoking tools, and performing actions within an enterprise environment.
This shift shifts the focus of security from just the output of the model to the behavior of the AI system in a live setting. The attack surface now includes agent workflows, delegated actions, non-human access, and shadow agents operating within business systems.
The platform leverages Check Point’s existing AI security platform, as well as technology from the ThreatCloud AI, Lakera and Cyata acquisitions. Combine discovery, governance, observability, runtime control, and continuous validation across the entire AI execution lifecycle.
At the heart of the system is what Check Point calls an AI-native security engine. Make real-time decisions using analysis of millions of AI interactions, adversarial testing, and threat intelligence, delivering sub-50ms response times in 100+ languages.
3 modules
This release includes three main modules, each of which will be available at different stages.
Workforce AI Security is available out-of-the-box and focuses on how employees use AI-powered applications. Provide visibility, governance, and runtime protection to enforce policies across approved and unapproved AI tools in real-time.
AI applications and agent security are also readily available. It’s designed to discover where AI is being used across your organization, evaluate the data and tools those systems have access to, evaluate their behavior, and manage the permissions and trust relationships associated with agent operations.
AI Red Teaming is a limited release. This part of the platform provides continuous adversarial testing of prompts, inference paths, workflows, tool usage, and agent behavior to identify weaknesses before the system is deployed more broadly.
This announcement reflects a broader industry effort to secure AI systems in production, rather than relying solely on development-stage controls and model guardrails. In practice, this means monitoring and restricting how AI applications and agents behave when linked to internal systems, sensitive data, and business processes.
David Haber outlined the company’s changing view of the market. “Enterprises are entering the agent era. AI is no longer limited to content generation. AI is accessing systems, using tools, chaining actions, and starting to operate with increased autonomy. This is changing the security model,” said David Haber, VP of AI Security at Check Point.
He added, “The challenge is no longer about what AI says, but what AI can do. Organizations need more than security in their models. They need runtime control over how AI behaves in the real world. AI Defense Plane provides that control across employees, applications, and AI agents.”
Check Point argues that governance and enforcement must occur at runtime, and business risks arise when AI interacts with operational systems. This perspective is gaining traction as companies test software agents that can query infrastructure, trigger workflows, and manipulate internal data without direct human intervention at each stage.
Red team elements also attracted support from outside Checkpoint. said George Davis, product lead at Sierra. “As AI can query infrastructure, trigger workflows, and manipulate sensitive data, the risks are no longer theoretical. Organizations need continuous testing to understand how these systems operate, where controls fail, and how resilient they are in production.”
AI Defense Plane is part of Check Point’s extensive AI security portfolio. The Workforce and Application modules are currently available, but the Red Team module remains in limited release.
