Ex-Meta execs raise $180M at $1.2B valuation to stop AI agents from trashing your endpoints

Glow, a startup founded by former Meta and Snowflake executives, just pulled back the curtain on a $180 million Series A that values the company at $1.2 billion. Sequoia Capital, Cyberstarts, Greenoaks, and Redpoint Ventures led the round. The unicorn status is notable because Glow is pre-revenue-metrics — a big bet that the way we secure laptops and servers needs a ground-up rewrite for the age of AI agents.

CEO Roi Tiger, who previously ran engineering at Meta, put it bluntly: “Suddenly, AI lands on the endpoint in a way we’ve never seen.” His platform uses AI agents from Anthropic and Google’s Gemini (via Amazon Bedrock) to map enterprise environments, assess risk in real time, and enforce policies. The goal isn’t just detection. Glow wants to block risky software and AI tools before they ever touch a device.

The timing isn’t accidental. Concerns about AI-powered attacks have spiked since Anthropic showed its Mythos model finding and exploiting software vulnerabilities with worrying skill. Glow says it’s already stopped malicious npm packages, flagged AI agents pulling in sketchy dependencies, and spotted devices where existing security tools were broken or missing — all inside paying customer environments. The company won’t name those customers, but says deployments span tens of thousands of devices globally.

The endpoint security market is a knife fight. CrowdStrike, Microsoft, SentinelOne, and Palo Alto Networks are all swinging. Glow’s pitch is that those products were built to detect threats after they emerge, not to prevent a new class of AI-driven risk from getting a foothold. With 100 employees split between Israel and the U.S., the startup is small. Whether “AI-native endpoint security” becomes a real category or just a good fundraising slide remains the billion-dollar question.