Check Point Announces 'AI Defense Plane' to Integrally Protect Enterprise AI Utilization
Check Point Software Technologies has unveiled 'AI Defense Plane,' an integrated security foundation designed to visualize and control how AI is connected, deployed, and operated across a business. As AI evolves from simple assistants to autonomous agents capable of accessing data and executing actions, this platform provides essential security intelligence, including runtime controls, to manage risks in real-world AI workflows.
📋 Article Processing Timeline
- 📰 Published: March 28, 2026 at 00:57
- 🔍 Collected: March 28, 2026 at 21:59 (21h 1m after Published)
- 🤖 AI Analyzed: April 15, 2026 at 04:02 (414h 2m after Collected)
Check Point® Software Technologies Ltd. (NASDAQ: CHKP), a pioneer and global leader in cybersecurity solutions, announced the 'Check Point AI Defense Plane' on March 23, 2026 (US time). This is an integrated AI security foundation designed to provide visibility and control over how AI is connected, introduced, and operated throughout a business. As AI systems evolve from mere assistants into agents that autonomously access data, utilize tools, and execute actions, the AI Defense Plane provides the necessary security intelligence foundation for the era of agentic AI.
David Haber, VP of AI Security at Check Point, stated:
'Enterprises are entering the agent era. AI is no longer just generating content; it is evolving to access systems, use tools, execute sequences of actions, and increase autonomy in operations, which fundamentally changes the security model. The challenge is no longer just what AI generates, but what it can do. Organizations need mechanisms to control AI not only in terms of model safety, but also in terms of actual behavior in the field at runtime. The AI Defense Plane realizes this control across employees, applications, and AI agents.'
Built upon Check Point's AI security platform, the AI Defense Plane is enhanced by ThreatCloud AI technology and technologies from recent acquisitions of Lakera and Cyata. It provides integrated detection, governance, observability, runtime control, and continuous verification throughout the AI execution lifecycle.
At the core of the AI Defense Plane is Check Point's AI-native security engine. This real-time decision engine is based on the analysis of millions of AI interactions, adversarial testing, and live threat intelligence, forming a structure where security is continuously strengthened as AI systems evolve. It achieves adaptive protection in under 50 milliseconds across over 100 supported languages, enabling high-speed defense even as the automation of attacks accelerates.
While many approaches focus on model guardrails, Check Point provides protection for the actual behavior of AI in production environments. The AI Defense Plane is designed to implement controls in scenarios where enterprise AI risks manifest, such as runtime, live environments, and workflows where AI and business operations collaborate.
Three Major Modules Comprising the AI Defense Plane:
- Workforce AI Security: Provides visibility, governance, and runtime safeguards for how employees utilize AI-enabled applications. It reduces the risk of sensitive data leakage through real-time policy enforcement and enables secure utilization across all AI tools, whether approved or unapproved.
- AI Application & Agent Security: Provides discovery, security posture management, and runtime control for AI applications and agentic systems integrated throughout the business. Organizations can identify where AI exists, understand the data and tools accessible, evaluate behavior, and manage the permissions and trust relationships that govern agentic AI execution.
- AI Red Teaming: Conducts continuous adversarial testing on prompts, inference paths, workflows, tool usage, and agent behavior. It helps organizations discover exploitable vulnerabilities early and strengthen resilience as AI systems transition from prototype to production.
These insights are supported by the latest research from Check Point Research (CPR). According to CPR's study, AI is causing a clear shift in the cyber threat landscape. As agentic workflows, delegated actions, and non-human access are incorporated into real business environments, attackers are moving from prompt-based exploitation to attacks targeting the AI agent architecture itself. For example, jailbreak techniques against Claude Code make it possible to bypass security features by tampering with configuration files.
Furthermore, according to the latest analysis in CPR's study on enterprise generative AI usage:
- 1 in 31 prompts carries a risk of sensitive data leakage (3.2% total)
- High-risk prompt activity confirmed in 90% of organizations using generative AI tools
- Potentially highly sensitive information, such as source code or confidential business data, present in 16% of all prompts
- Employees use an average of approximately 10 different generative tools and generate 69 prompts per month
As companies rapidly adopt AI, risks increase alongside usage, making visibility, governance, and guardrails essential.
George Davis, Product Leader at Sierra, stated:
'For agentic systems, red teaming is essential. When AI can query infrastructure, trigger workflows, and manipulate sensitive data, the risks are no longer theoretical. Through continuous testing, organizations need to understand how these systems can be manipulated, where controls might fail, and how resilient they are in production.'
Availability and RSA Conference Details:
- AI Defense Plane: Available within Check Point's AI security portfolio
- Workforce AI Security: Available simultaneously with this announcement
- AI Application & Agent Security: Available simultaneously with this announcement
- AI Red Teaming: Available to select customers
At RSA Conference 2026, held in San Francisco, USA, from March 23 to 26, 2026 (local time), the AI Defense Plane was exhibited, with live demonstrations and expert briefings conducted. Additionally, the hands-on exhibit 'Gandalf: The Agent Gauntlet' was unveiled for the first time, allowing attendees to experience how agentic systems are attacked and manipulated, and to verify them using the latest red teaming techniques.
For more details on Check Point's AI Defense Plane, please visit here: https://www.checkpoint.com/jp/ai-security/
This press release was created based on the press release announced on March 23, 2026 (US time).
David Haber, VP of AI Security at Check Point, stated:
'Enterprises are entering the agent era. AI is no longer just generating content; it is evolving to access systems, use tools, execute sequences of actions, and increase autonomy in operations, which fundamentally changes the security model. The challenge is no longer just what AI generates, but what it can do. Organizations need mechanisms to control AI not only in terms of model safety, but also in terms of actual behavior in the field at runtime. The AI Defense Plane realizes this control across employees, applications, and AI agents.'
Built upon Check Point's AI security platform, the AI Defense Plane is enhanced by ThreatCloud AI technology and technologies from recent acquisitions of Lakera and Cyata. It provides integrated detection, governance, observability, runtime control, and continuous verification throughout the AI execution lifecycle.
At the core of the AI Defense Plane is Check Point's AI-native security engine. This real-time decision engine is based on the analysis of millions of AI interactions, adversarial testing, and live threat intelligence, forming a structure where security is continuously strengthened as AI systems evolve. It achieves adaptive protection in under 50 milliseconds across over 100 supported languages, enabling high-speed defense even as the automation of attacks accelerates.
While many approaches focus on model guardrails, Check Point provides protection for the actual behavior of AI in production environments. The AI Defense Plane is designed to implement controls in scenarios where enterprise AI risks manifest, such as runtime, live environments, and workflows where AI and business operations collaborate.
Three Major Modules Comprising the AI Defense Plane:
- Workforce AI Security: Provides visibility, governance, and runtime safeguards for how employees utilize AI-enabled applications. It reduces the risk of sensitive data leakage through real-time policy enforcement and enables secure utilization across all AI tools, whether approved or unapproved.
- AI Application & Agent Security: Provides discovery, security posture management, and runtime control for AI applications and agentic systems integrated throughout the business. Organizations can identify where AI exists, understand the data and tools accessible, evaluate behavior, and manage the permissions and trust relationships that govern agentic AI execution.
- AI Red Teaming: Conducts continuous adversarial testing on prompts, inference paths, workflows, tool usage, and agent behavior. It helps organizations discover exploitable vulnerabilities early and strengthen resilience as AI systems transition from prototype to production.
These insights are supported by the latest research from Check Point Research (CPR). According to CPR's study, AI is causing a clear shift in the cyber threat landscape. As agentic workflows, delegated actions, and non-human access are incorporated into real business environments, attackers are moving from prompt-based exploitation to attacks targeting the AI agent architecture itself. For example, jailbreak techniques against Claude Code make it possible to bypass security features by tampering with configuration files.
Furthermore, according to the latest analysis in CPR's study on enterprise generative AI usage:
- 1 in 31 prompts carries a risk of sensitive data leakage (3.2% total)
- High-risk prompt activity confirmed in 90% of organizations using generative AI tools
- Potentially highly sensitive information, such as source code or confidential business data, present in 16% of all prompts
- Employees use an average of approximately 10 different generative tools and generate 69 prompts per month
As companies rapidly adopt AI, risks increase alongside usage, making visibility, governance, and guardrails essential.
George Davis, Product Leader at Sierra, stated:
'For agentic systems, red teaming is essential. When AI can query infrastructure, trigger workflows, and manipulate sensitive data, the risks are no longer theoretical. Through continuous testing, organizations need to understand how these systems can be manipulated, where controls might fail, and how resilient they are in production.'
Availability and RSA Conference Details:
- AI Defense Plane: Available within Check Point's AI security portfolio
- Workforce AI Security: Available simultaneously with this announcement
- AI Application & Agent Security: Available simultaneously with this announcement
- AI Red Teaming: Available to select customers
At RSA Conference 2026, held in San Francisco, USA, from March 23 to 26, 2026 (local time), the AI Defense Plane was exhibited, with live demonstrations and expert briefings conducted. Additionally, the hands-on exhibit 'Gandalf: The Agent Gauntlet' was unveiled for the first time, allowing attendees to experience how agentic systems are attacked and manipulated, and to verify them using the latest red teaming techniques.
For more details on Check Point's AI Defense Plane, please visit here: https://www.checkpoint.com/jp/ai-security/
This press release was created based on the press release announced on March 23, 2026 (US time).