Malaysian companies are rolling out AI assistants, agents and chatbots faster than their security teams can keep track of them, and most of that activity travels straight across the corporate network, where older firewalls were never built to inspect it. Check Point's answer is to turn the firewall a company already runs into the thing that watches its AI traffic.

Check Point Software Technologies (NASDAQ: CHKP) has released the Check Point AI Network Firewall, delivered as part of its firewall software update R82.20 and available now. Rather than asking businesses to buy and deploy a separate appliance, it adds AI traffic inspection to the physical or virtual Check Point firewalls they already operate, with no new infrastructure and no need to redesign the network.

Editor
Features Editor

Dinesh Raj chevron_right

Features and explainers on how technology is changing the way Malaysians work, shop and create.

Table of Contents

What the Check Point AI Network Firewall actually does

The firewall focuses on three kinds of AI activity that traditional network security tends to miss:

  • Employee AI use: it discovers the AI apps, agents and tools staff are using, both sanctioned and unofficial, shows how they are being used, and can stop sensitive company data from leaving the network based on what a prompt is trying to do. Check Point Research reports that organisations now run around ten AI applications a month, many outside any formal process.
  • AI tools (MCP): it detects Model Context Protocol traffic, the emerging standard that lets AI agents call external tools, and lets teams set rules for what those agents can reach. Check Point Research found security weaknesses in 40% of 10,000 MCP servers it reviewed.
  • AI apps and language models: it blocks prompt injection and malicious inputs before they reach a model, working inline without changes to the application. The company's researchers identified 15,300 hidden injection payloads planted in public web pages, about 70% of them in parts of the page no person ever sees.

Those figures point to a wider problem. Check Point's AI Security Report 2026 found that between 87% and 93% of organisations hit at least one high-risk generative-AI interaction every month, and the share of prompts carrying sensitive corporate, personal or regulated data doubled in a year to one in every 25.

Built into the firewall, not bolted on

The pitch is about consolidation. Many businesses add separate AI security products alongside their existing defences, building up a pile of disconnected tools. Check Point is instead folding the capability into its AI Defense Plane, a single control layer that also covers endpoints, cloud, applications and APIs, and extends central policy management to its SASE and SD-WAN products.

"By bringing AI security into the firewall organizations already run, we give security teams the visibility and control to enable AI adoption safely, in real time," said Nataly Kremer, Chief Product Officer at Check Point.

Pete Finalle, a research manager at analyst firm IDC, noted that building AI security into existing enforcement points, rather than adding yet another standalone tool, can improve visibility and simplify management, though he described the native approach as still uncommon.

Availability

The Check Point AI Network Firewall is available now for existing Check Point firewall customers as part of the R82.20 software release. Check Point has not published separate pricing for the feature. Full details are on the company's AI Network Firewall page.

For any organisation already worried about staff pasting confidential information into a chatbot, the useful takeaway is the direction of travel: AI traffic is becoming something the network itself is expected to police, not a separate problem handled by a separate box.