Clalit Health Services, Israel’s largest health maintenance organization, has signed a multimillion-shekel strategic agreement with Check Point Software Technologies to build a security framework for its use of artificial intelligence, CTech by Calcalist reported Sunday. The centerpiece is a gateway meant to stop sensitive patient information from flowing into public AI services.

What Happened

Clalit serves about five million people. Under the deal, Check Point will develop a security gateway that sits between Clalit’s staff and outside AI tools. It is designed to prevent uncontrolled use of public chatbots such as ChatGPT and Gemini and to block personally identifiable information from being sent to external platforms.

The system will also enforce privacy rules and guardrails for approved AI use, and monitor and block anomalous behavior by external AI agents and third-party applications in real time.

RelatedJPMorgan Bars Outside AI Agents From Customer Accounts as Dimon Warns of 'Rogue Agents'

Clalit’s chief information security officer, Avivit Kotler, told CTech that Check Point will see only metadata. She said the vendor will have no access to patient data.

Full deployment is targeted for the first quarter of 2027. Check Point will retain the intellectual property and intends to commercialize the platform for other customers.

Why It Matters

The deal addresses one of the most common ways generative AI creates risk inside large organizations: employees copying confidential material into consumer chatbots to summarize or rewrite it. In a health system, that material can include diagnoses and personal identifiers covered by strict privacy rules.

It also reflects a newer concern. As AI agents and third-party apps gain the ability to take actions on their own, security teams need controls that judge what those tools do, not only what data users type. Building that into a gateway for a five-million-member insurer and provider gives Check Point a large reference deployment for a product category that many security vendors are racing to fill.

Under Sustained Attack

Clalit is operating under heavy pressure from state-linked hackers. Kotler said that in recent years the organization has faced attack attempts from Iran and its proxies at rates hundreds of times higher than usual, which she put at tens of thousands of attacks a year.

Israeli healthcare has been hit before. CTech noted that the ransomware attack on Hillel Yaffe Medical Center led the Health Ministry to tighten cybersecurity regulations for the sector.

Clalit already uses AI in controlled settings. It runs Aidoc’s medical imaging algorithms and its own models that analyze historical data to identify disease, both in closed, isolated environments. The new gateway is aimed at the less controlled side of AI adoption, where staff reach for public tools.

What to Watch

The rollout through the first quarter of 2027 will show whether the gateway can filter AI traffic without slowing clinical work. Check Point’s plan to sell the platform commercially means the Clalit deployment is likely to serve as its showcase, so watch for a product launch and other healthcare or regulated-industry customers.