Glow has come out of stealth with $180 million and a blunt claim: endpoint security built for human users won't be enough when AI agents start acting on corporate devices.
Glow launched publicly on July 22, 2026, with $180 million in funding and a $1.2 billion valuation, making it a cybersecurity unicorn before most buyers had seen the product in the wild. According to the company's announcement distributed by GlobeNewswire, Sequoia Capital, Cyberstarts, Greenoaks, and Redpoint Ventures led the round, with Index Ventures, Swish Ventures, Lux Capital, Operator Collective, and Holly Ventures also participating.
The pitch is direct. CrowdStrike, SentinelOne, Microsoft Defender, and the rest of the endpoint detection market were built for a corporate device where the main actor was a person. Now a developer can plug GitHub Copilot or Cursor into an IDE, a sales team can connect an agent to a CRM, and a security team has to track software that reads files, calls APIs, and moves through workflows at machine speed. That's a different problem.
Glow was founded in Israel in 2025 by Roi Tiger, Omer Singer, and Ophir Arie. Tiger spent nine years at Meta and was most recently a VP of Engineering. Singer was Snowflake's Head of Cybersecurity Strategy. Arie was VP of R&D at Claroty. The broader team includes Chief Product Officer Arnon Joseph, another Meta veteran, and COO Emily Heath, who served as CISO at United Airlines and DocuSign and was on Wiz's board through its $32 billion sale to Google.
That is not a team assembled to sell a small add-on. It's a team built to take a run at a category.
The bet is prevention - not more alerts
Glow says its platform uses specialised AI agents to map enterprise environments, analyse endpoint risk in real time, and enforce policies automatically. The company frames the product around three jobs: asset intelligence, software control, and safe AI adoption. The practical claim is simple enough. Security teams shouldn't have to manually review every new AI tool a developer downloads on a Tuesday afternoon.
Glow's launch materials say regular usage of AI on corporate devices, authorized or not, has climbed from 15% to 45% in one year. That figure matters because you can't secure what you don't see. If AI tools are already running across employee laptops, the old cycle of waiting for suspicious behavior and then investigating it starts to look painfully slow.
Roi Tiger put the company's view plainly in the launch announcement, saying prevention had always been the right answer in security but had not worked at enterprise scale without blocking the business. Emily Heath made the CISO version of the same point, arguing that enterprises will adopt AI and the real question is whether security can keep up. Those are vendor lines, of course. But they are pointed at a real gap.
Early customers span healthcare, retail, and financial services, according to Glow, though the company hasn't named them publicly. Keep that caveat in view. A $1.2 billion valuation before a broad public product rollout tells you more about investor demand than customer proof. Frankly, both things matter.
The incumbents are not asleep
CrowdStrike is already moving into this territory. In March 2026, the company announced Falcon capabilities for AI agent discovery, shadow AI governance, and runtime threat detection across endpoints, SaaS, browser, and cloud environments. CrowdStrike also said its sensors detected more than 1,800 distinct AI applications running on enterprise devices, representing nearly 160 million unique application instances across its customer base.
SentinelOne is pushing in the same direction. Its February 2026 product announcement framed AI security around data protection, infrastructure hardening, and runtime defence - and later messaging focused on human and non-human identities as autonomous agents spread through business systems. Microsoft has the distribution advantage with Defender and Copilot sitting inside the same enterprise accounts. Nobody should pretend this market is empty.
Glow's argument is sharper than that. It is saying retrofitting AI visibility onto older endpoint platforms won't be enough because the agent itself has become a first-class actor on the device. You don't have to buy that argument completely to see why investors did. Every large company now wants more AI automation, but the security team is still accountable when an agent touches the wrong file, leaks a credential, or sends data to a service nobody approved.
The money will go toward building Glow's go-to-market team in the United States and expanding Glow Labs, its research arm. The company also says it will show the product at Black Hat USA 2026 in Las Vegas, with a House of Glow event at the House of Blues Courtyard from August 3 through August 5.
That public showing will matter more than the launch language. A unicorn valuation gets attention. Security buyers will want something duller and harder: proof that Glow can see real endpoint activity, make fast policy decisions, and avoid becoming one more console asking an already stretched team to look at more alerts.
Also read: Alphabet reports Q2 2026 earnings today with its $190 billion AI bet under the microscope • Computer science enrollment falls for the first time in 20 years as AI reshapes who gets hired • CuspAI raises $450 million to let AI design the next generation of chip materials
