Israeli cybersecurity firm Check Point buys Deepchecks

Israel-based cybersecurity company Check Point has acquired the team and intellectual property of Deepchecks to expand its AI security tools.

Industry sources valued the deal at US$10 million to US$20 million.

Deepchecks was founded in 2019 by CEO Philip Tannor and CTO Shir Chorev.

The company builds software to evaluate and monitor AI systems and has raised US$14 million, according to PitchBook.

Check Point said the deal will support its new Agentic Network Security Orchestration platform that uses AI agents to manage firewall policies and access controls and help with troubleshooting and compliance.

The acquisition is Check Point’s fourth Israeli cybersecurity deal this year after Cyclops Security, Cyata, and Rotate.

The company faces pressure to strengthen its AI and cloud security business after first-quarter revenue rose 5% to US$668 million.

🔗 Source: Calcalist

🧠 Food for thought

Implications, context, and why it matters.

Check Point wants to build trust in AI-run network security

Check Point built the platform for networks that many IT and security teams can no longer manage by hand. Hybrid cloud growth plus mergers and acquisitions have added to that strain 1.
Its agents rely on a “Network Knowledge Graph,” not generic AI. Check Point calls it a live model of a customer’s environment that updates with network layout, traffic flows, asset dependencies, plus real-time configuration data 2.
The Deepchecks deal adds a layer that measures performance over time and helps improve the system. Check Point said it signed a definitive agreement to acquire the team plus intellectual property of Deepchecks, rather than the company itself 1.
Security teams still control intended actions. They can approve high-impact changes before execution and review an execution trace, a record of what the system did plus why 1.

Check Point is pushing past AI help toward automated action

Check Point wants to move from AI assistance to autonomous agents that carry out network security tasks within preset limits, policies, plus human oversight 1.
The company says that level of autonomy depends on decisions tied to a customer’s live network environment, not static training data 1.
Check Point treats reasoning over a live proprietary data model as the difference between its approach and more generic AI for security 2.