Glow emerges from stealth at $1.2B valuation to challenge endpoint security in the AI era

Unicorn status before you've even shown revenue numbers used to be rare in cybersecurity. Now it's basically the entry fee. Glow, founded this year by a mix of Meta, Snowflake, and Claroty alumni, just pulled off exactly that trick, landing a $180 million all-equity Series A at a $1.2 billion valuation, with Sequoia, Cyberstarts, Greenoaks, and Redpoint leading the round.

The pitch is that endpoint security, the business of protecting laptops, servers, and everything else employees touch, is about to get scrambled by AI in ways CrowdStrike, Microsoft, SentinelOne, and Palo Alto Networks weren't built for. CEO Roi Tiger frames it simply: the last decade was about everything moving to the cloud. This decade, AI is landing directly on the endpoint, running as agents and developer tools on machines that legacy detection systems were never designed to watch this closely. Anthropic's Mythos model, which reportedly showed real skill at finding and exploiting software flaws, has only sharpened the anxiety driving that pitch.

What Glow actually built is a platform that uses AI agents of its own to continuously map what's running across an organization's devices, score the risk, and block bad stuff before it lands rather than cleaning up after. Tiger says the system has already caught malicious npm packages before install, flagged AI agents trying to pull in sketchy code, and spotted devices where standard detection tools had quietly stopped working properly. That last one is the kind of unglamorous gap that keeps CISOs awake, and it's a smart wedge into a market where the big players are optimized for after-the-fact alerts.

Glow isn't building its own foundation models. It leans on Anthropic and Google's Gemini through Amazon Bedrock, then wraps that with proprietary software meant to feed enterprise context into those models and make them dependable enough for security decisions. That's a pragmatic choice, and it mirrors how a lot of AI-security startups are choosing to compete on integration and context rather than raw model muscle.

The company says it already has paying customers in healthcare, retail, and financial services, with deployments running into tens of thousands of devices, though it won't name names. With roughly 100 employees, 70 percent of them in Israel, Glow is small next to the incumbents it's challenging. Whether "AI-native endpoint security" becomes its own category, or just a feature the big vendors eventually bolt on, is genuinely unsettled. Enterprises are still figuring out what AI agents on employee devices even mean for their risk posture.