Check Point Targets New Traffic Threats With 'AI Network Firewall'
TECHWORLD ·
✦ AI Summary
Check Point Software Technologies announced on the 15th that it had launched the "AI Network Firewall."
The product is part of the firewall software R82.20 release and is delivered through the "AI Defense Plane" without requiring additional infrastructure or architecture changes.
Check Point said the firewall supports identifying AI apps, agents, and tools, preventing sensitive data leakage, detecting MCP communications, and blocking prompt injection.
Check Point Software Technologies (Check Point) announced on the 15th that it had launched the "AI Network Firewall." The "AI Network Firewall" is a product that integrates AI security features into a traditional firewall, and Check Point said it was introduced to address new traffic threats.
The "AI Network Firewall" is available as part of Check Point's firewall software R82.20 release. The feature requires no additional infrastructure or architecture changes, and is delivered through Check Point's "AI Defense Plane."
The "AI Defense Plane" is an architecture that supports AI detection, management, and protection. It supports local AI agent detection and control as well as SaaS AI agent detection and control, and also provides runtime protection and governance for AI applications. It also supports risk detection and guardrails for self-developed and deployed AI protection.
The AI Network Firewall is a component of that architecture and provides an integrated control plane for AI detection, control, and protection across networks, endpoints, cloud, applications, and API. The firewall is presented as a unified control tool for handling AI in one place across multiple environments and touchpoints.
Through this, companies can use the AI Network Firewall to build a protection framework for employee AI use. They can also build protection frameworks for AI tools (MCP), AI applications, and LLMs.
The AI Network Firewall identifies the AI apps, agents, and tools in use, and determines how AI is being used, the use cases involved, and the intent, including both official and private use. It also prevents sensitive data from leaking outside the network in management and prompt-related use cases, while enforcing access control policies across all interactions.
The firewall detects Model Context Protocol (MCP) communications and improves visibility into servers and tools in use. It then prevents prompt injection, blocks malicious inputs, and stops harmful prompts before they reach the LLM.
The company said that the vulnerabilities of traditional firewalls lie in new AI-driven traffic attacks such as prompts, autonomous agent tasks, and sensitive business context, and emphasized the possibility that this solution can complement the weaknesses of existing firewalls.
According to Check Point Research's "AI Security Report 2026," 87% to 93% of companies experienced at least one high-risk generative AI interaction per month. The share of prompts containing sensitive corporate, personal, or regulated data doubled in one year, and such prompts accounted for about 4% of all interactions, or 1 in 25. The report analyzed that the pace of AI adoption at companies is outstripping manageable speeds, and that activities requiring management have already spread across the network.
Nataly Kremer, CPO at Check Point, said that as AI-driven innovation in enterprise networks advances, every prompt, model call, and agent interaction is already concentrated on the network. She explained that traditional firewalls were not designed to detect and control such AI activity, and said Check Point aims to integrate AI security capabilities into the firewalls companies already operate, providing security teams with visibility and control to support real-time AI adoption and help enable safe AI adoption.
Source: TECHWORLD · Kim Hye-jin
Original: https://www.epnc.co.kr/news/articleView.html?idxno=406937
References
This article was produced with the help of an automated content generation algorithm.
Source: TECHWORLD
View originalThis article was summarized and organized by BizCrush based on the original article from TECHWORLD. For exact quotations and full details, please refer to the original article.