Secretive Glow targets endpoint risks from enterprise AI agents

Glow is raising more than $100 million at a valuation above $1 billion before unveiling a public product, betting that enterprise AI agents and developer tools have opened a new class of endpoint risk. Startup Nation Central’s Finder database lists Glow Technology as an enterprise endpoint protection company, placing the secretive Israeli startup squarely inside a crowded security category.

The company is tied to Roi T, identified as an Onavo co-founder, a background that gives Glow credibility in Israeli cyber circles but leaves open a basic question for buyers: what, exactly, does it protect that existing endpoint tools do not? The pitch lands as companies add autonomous software to internal workflows, where AI agents can move quickly, call tools, and operate with broad privileges across code, data and administration systems.

Investors have already begun funding the category around Glow’s thesis. Torq raised $140 million at a $1.2 billion valuation and described itself as the “Agentic AI security operations pioneer,” saying customer adoption of AI agents is fueling its growth. Bold Security emerged from stealth with $40 million to build an AI security agent for endpoints, underscoring how quickly capital is chasing products built for machine-speed operations inside companies.

The broader industry is treating agents as a security boundary, not just a feature. Menlo Ventures says AI has created “a new era of cybersecurity,” and its security practice says it is backing founders with deep cybersecurity expertise as new categories emerge. CyberArk says AI agents create “a new identity class” and an “expanded attack surface,” while Box has framed enterprise trust for AI agents as a scale problem. The Cloud Security Alliance and Zenity have argued that visibility, ownership and control have not kept pace as agents reach enterprise scale.

That context makes Glow’s timing clear, but it also sharpens the test. Security buyers already understand endpoint protection for laptops, phones and servers. Glow will have to prove that AI agents and developer tools create a distinct enough risk surface to justify a fresh category, rather than another repackaging of endpoint security with an AI label. The valuation implies conviction; the market will need substance.