A cybersecurity startup that has never shipped a public product just became a billion-dollar company. Glow, an Israeli firm founded in 2025 by Roi Tiger, stepped out of stealth in July 2026 carrying $180 million in fresh funding and a valuation exceeding $1 billion — built entirely on the argument that AI endpoint security, as enterprises currently practice it, is dangerously broken.
Summary
Key takeaways
- Glow is an Israeli cybersecurity startup founded in 2025 by Roi Tiger, valued at over $1 billion before launching any public product.
- The company raised $180 million from Sequoia Capital, Index Ventures, Cyberstarts, Greenoaks, and Redpoint.
- Research cited by Glow shows 30% of enterprise AI agents operate without adequate controls; 67% of software in the average organization goes entirely unmonitored.
- Glow’s platform uses AI agents for asset intelligence, software control, and autonomous remediation of security threats.
- All performance metrics — including a claimed 90% reduction in AI supply-chain risk — are self-reported and have not been independently validated.
Glow’s founding and the bet investors are making
The investor roster backing Glow reads like a who’s who of tier-one venture capital: Sequoia Capital, Index Ventures, Cyberstarts, Greenoaks, and Redpoint all participated. The company had quietly surfaced as early as February 2026, when initial reports indicated it had already crossed the $100 million funding mark at a billion-dollar-plus valuation while remaining almost entirely in stealth mode. The July 2026 emergence confirmed the full picture.
What makes this raise unusual isn’t the dollar figure — it’s the timing. Glow reached unicorn status without a publicly available product, which reflects less about the company’s traction and more about how urgently the security industry believes the AI agent problem needs solving. Investors are essentially pricing in the threat before the solution has been proven at scale.
What the platform actually does
Glow’s core argument is straightforward: enterprise security was architected for humans. Firewalls, endpoint detection, and access controls all assumed that a person was behind the keyboard making decisions. AI agents — autonomous software that can write code, query APIs, and trigger deployments — broke that assumption entirely.
Glow’s platform deploys its own AI agents to counter that gap, operating across three functions: asset intelligence, which continuously maps everything running on an endpoint; software control, which governs what tools and packages are permitted; and what the company calls “safe AI adoption,” a policy-driven layer designed to ensure that AI agents operating inside an enterprise stay within defined boundaries.
Autonomous remediation and the 28-day resolution benchmark
The platform’s most commercially differentiated feature is autonomous remediation. Rather than alerting a security team and waiting for a human to triage and respond, Glow’s system acts on its own — fixing problems without requiring ticket approval. Early customer data, which the company reports internally, indicates an average resolution time of 28 days for critical issues, alongside a continuously updated inventory the firm describes as a “living asset inventory.”
The numbers behind the problem Glow is solving
The company points to research showing that 30% of AI agents in enterprise environments operate without adequate controls or guardrails. Separately, 67% of software in the average organization remains completely unmonitored. Those two figures together describe an attack surface that has grown faster than the security tooling built to cover it.
Glow also claims a 90% reduction in AI supply-chain risk for its initial users. That number is striking, but it warrants context: all of Glow’s performance metrics are self-reported from early customer interactions, and no independent benchmarks exist yet. The security community will get its first real opportunity to scrutinize the technology when Glow attends Black Hat USA 2026.
Why crypto and Web3 firms face the sharpest exposure
Crypto security threats have evolved well beyond phishing emails and exchange hacks. Crypto firms, DeFi protocols, and Web3 startups are among the most aggressive adopters of AI coding assistants and autonomous agents — often deploying them to audit smart contracts, automate trading strategies, and accelerate development cycles. That speed comes at a cost.
The Lazarus Group‘s social engineering campaigns against crypto developers, the steady proliferation of malicious npm packages targeting Web3 wallets, and the recurring wave of supply-chain attacks all exploit exactly the kind of visibility gaps Glow claims to close. An AI agent granted access to deployment keys without security guardrails is, in practical terms, a direct path to a protocol’s treasury. The threat isn’t hypothetical — it’s already playing out across the ecosystem, and the tooling to stop it has not kept pace.
Why existing security vendors aren’t enough
Established endpoint detection and response players — CrowdStrike, SentinelOne, and Microsoft Defender — were all designed and built before the AI agent era. Their architectures assume human-initiated behavior, making them structurally ill-suited to monitor or control autonomous agents that operate continuously, generate their own network traffic, and interact with systems in ways no human workflow anticipated.
The commercial signal here is worth noting. CrowdStrike was valued at roughly $6.6 billion at its 2019 IPO, after years of revenue generation and broad market validation. Glow’s $180 million raise at a comparable entry valuation — with no public product yet — captures just how much the market is willing to pay for a credible solution to a problem that the incumbents haven’t addressed.
That valuation gap also carries a warning. Glow will need to translate early customer results into verifiable, independently audited performance data. The Black Hat USA 2026 appearance will be the first real stress test — a venue where the security community’s scrutiny is both rigorous and public. How the platform holds up under that pressure will matter far more to its long-term positioning than any funding announcement.
FAQ
What is Glow and when was it founded?
Glow is an Israeli cybersecurity startup founded in 2025 by Roi Tiger, focused on AI-native endpoint protection. It emerged from stealth mode in July 2026.
How much funding has Glow raised and who are the investors?
Glow has raised $180 million from a group of investors that includes Sequoia Capital, Index Ventures, Cyberstarts, Greenoaks, and Redpoint.
What security challenges does Glow address?
Glow targets the security risks created by AI agents operating inside enterprise environments, offering asset intelligence, software control, safe AI adoption policies, and autonomous remediation of threats — without waiting for human intervention.
Are Glow’s effectiveness claims independently verified?
No. Glow’s performance metrics — including the claimed 90% reduction in AI supply-chain risk — are currently self-reported from early customer interactions and have not been independently validated. Black Hat USA 2026 will provide the first opportunity for broader community evaluation.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.
