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

Glow, a cybersecurity startup founded by erstwhile Meta and Snowflake executives, emerged from stealth arsenic a unicorn, betting that artificial quality is reshaping however enterprises unafraid worker devices.

The Palo Alto-headquartered startup connected Wednesday said it raised $180 cardinal successful an all-equity Series A backing circular that valued it astatine $1.2 billion, with backing from Sequoia Capital, Cyberstarts, Greenoaks, and Redpoint Ventures, alongside information from Index Ventures, Swish Ventures, Lux Capital, Operator Collective, and Holly Ventures. The concern made Glow 1 of the latest cybersecurity startups to execute unicorn presumption earlier publically disclosing gross metrics.

As much enterprises deploy AI tools and attackers progressively usage generative AI to automate phishing, make malware, and motorboat much blase cyberattacks, companies are rethinking however they unafraid endpoints — from worker laptops to servers and different connected devices. Concerns person intensified since Anthropic unveiled its Mythos AI model, which the institution said demonstrated precocious capabilities successful identifying and exploiting bundle vulnerabilities, prompting broader statement implicit AI-assisted cyberattacks. Glow is betting that this displacement requires a caller attack to endpoint security.

Founded successful 2025, Glow is gathering an endpoint information level that helps enterprises show and power the software, AI agents, and developer tools moving connected worker devices. The startup said the level uses specialized AI agents to continuously representation endeavor environments, measure hazard successful existent time, and enforce information policies.

“If you deliberation of the past decade, everything was moving to the unreality and SaaS. Suddenly, AI lands connected the endpoint successful a mode we’ve ne'er seen,” co-founder and main enforcement Roi Tiger (pictured above, center) said successful an interview.

Tiger, a erstwhile Meta vice president of engineering, co-founded Glow alongside erstwhile Snowflake cybersecurity strategy caput Omer Singer (pictured above, left), erstwhile Claroty vice president of probe and improvement Ophir Arie (pictured above, right), and erstwhile Meta engineering person Arnon Joseph. The startup’s enactment squad besides includes main operating serviceman Emily Heath, a erstwhile main accusation information serviceman astatine United Airlines and DocuSign who served connected the committee of Wiz done its $32 cardinal acquisition by Google and was antecedently a spouse astatine Cyberstarts.

Even though it has lone conscionable emerged from stealth, Glow said it already has paying customers crossed industries including healthcare, retail, and fiscal services. However, it declined to disclose lawsuit names and numbers. The startup’s emblematic deployments, Tiger said, span tens of thousands of worker devices crossed planetary organizations.

To powerfulness the platform, Glow uses AI models from Anthropic and Google’s Gemini done Amazon Bedrock, portion gathering its ain bundle to supply the models with endeavor discourse and amended their reliability for information tasks, Tiger told TechCrunch.

Glow’s platform, Tiger said, has already prevented malicious npm packages, third-party bundle components utilized to physique applications, from being installed successful lawsuit environments, identified AI agents attempting to propulsion successful specified software, and detected worker devices wherever endpoint detection and effect tools were missing oregon operating with reduced functionality.

Glow enters a crowded endpoint information marketplace dominated by companies including CrowdStrike, Microsoft, SentinelOne, and Palo Alto Networks. Tiger said existing endpoint detection and effect products absorption chiefly connected detecting threats aft they emerge, whereas Glow is designed to forestall risky software, AI agents, and developer tools from entering endeavor environments successful the archetypal place.

The startup employs astir 100 people, astir 70% of whom are successful Israel and the remainder successful the U.S. Whether AI-native endpoint information platforms go a chiseled class remains to beryllium seen, arsenic enterprises are lone opening to grapple with the information implications of progressively susceptible AI models.