LLM evaluation capability is becoming a cybersecurity acquisition target.
Check Point Software announced a definitive agreement to acquire Deepchecks, an AI testing and LLM evaluation company founded in Israel, on May 19, 2026. The acquisition brings Deepchecks’ team and IP into Check Point’s Agentic Network Security Orchestration platform, a product focused on applying AI to network security monitoring and response.
The deal price is estimated by industry analysts at $10 million to $20 million, according to CTech and PitchBook data. That’s an estimate, not a confirmed figure, Check Point has not disclosed financial terms, and the estimate carries normal caveats for private M&A valuation data from secondary sources. Deepchecks raised $14 million since its 2019 founding, backed by Alpha Wave, Hetz, and Grove Ventures, according to the Wire’s background research.
The strategic rationale, per Check Point’s VP of AI Technologies, is integrating LLM evaluation into agentic security workflows. The logic is defensible: agentic AI systems are increasingly deployed in security operations, and the failure modes of agentic AI (hallucination, prompt injection, tool misuse) are themselves security vulnerabilities. A security vendor that can test and evaluate its own AI-driven products adds a credibility layer that pure-play security companies without evaluation capability can’t easily replicate.
The real story is the count. Per CTech, this is Check Point’s fourth Israeli cybersecurity startup acquisition in 2026. That cadence, four deals in roughly five months, is a deliberate consolidation strategy, not opportunistic deal flow. Check Point is building by acquisition in a specific geography (Israel’s cybersecurity ecosystem) and in a specific direction (AI-native security capabilities). The cumulative effect of four acquisitions is a product suite that would take years to develop organically.
This also fits the 48-hour acquisition pattern visible across this reporting cycle. ADI is buying the power delivery layer. Mistral is buying industrial simulation expertise. Check Point is buying LLM evaluation. Three different buyers, three different target capabilities, all in the same 24-hour window, each plugging a different gap in their AI stack. The pattern isn’t coincidence; it reflects a market where AI-native capability gaps are now priced and acquirable.
The Deepchecks deal size, estimated at $10 million to $20 million, puts it in the acqui-hire range. For a company that raised $14 million over seven years, the return is modest. The investors (Alpha Wave, Hetz, Grove) are likely recovering capital rather than capturing significant upside. This is a capability acquisition, not a financial event.
Analysis
Check Point's four acquisitions in five months share a common thread: each adds AI-native capability that traditional cybersecurity product development can't produce at speed. LLM evaluation (Deepchecks) joins a growing list of AI capability gaps being closed through M&A rather than R&D. The consolidation bet is that owning these capabilities early is worth the acqui-hire premium even at modest deal sizes.
What to watch
How Check Point’s Agentic Network Security Orchestration platform performs in enterprise procurement cycles over the next two quarters, specifically whether the Deepchecks integration becomes a differentiator in competitive evaluations against CrowdStrike, Palo Alto Networks, and SentinelOne, all of which are building their own AI evaluation and monitoring capabilities.
The Q3 cybersecurity vendor benchmark reports will be the first independent read on whether the Deepchecks IP meaningfully advances Check Point’s agentic security product.