Recommended Articles
VC & Investors

AI Security Is the Hottest Vertical in Venture Right Now

By Angeled · November 11, 2026 · VC & Investors
AI Security Is the Hottest Vertical in Venture Right Now
AI Security Is the Hottest Vertical in Venture Right Now

Few verticals in venture capital are moving as fast as AI security. In a single stretch of reporting, three companies raised substantial rounds: Daylight Security closed an $80 million Series B backed by Goldman Growth Equity and Bain, with Assaf Rappaport involved; Bay raised a $50 million Series A from Battery, on top of a $13 million seed; and Lumbra raised a $25 million Series A from Point72 and Ansa, with customers including Pentagon and intelligence community networks. In the same period, Fireworks AI disclosed that it had suffered multiple cyberattacks exposing customer data and unfinished models, on November 10, 2026.

The financings and the breach belong in the same story. They are two sides of one fact: artificial intelligence creates new attack surfaces at the same time as it creates new defences, and the market for both is expanding rapidly.

Why AI Creates New Attack Surfaces

Every AI system is a new thing to attack. Models can be tricked into revealing their training data. Agents with access to company systems can be manipulated into taking unauthorised actions. Unfinished models, like those Fireworks AI said were exposed, are valuable intellectual property in their own right. Customer data flowing through AI pipelines creates new places for it to leak.

The Fireworks AI disclosure is a useful case study. The company is an AI infrastructure provider — the kind of firm whose customers trust it with sensitive data and valuable models. That it suffered multiple attacks, and that customer data and unfinished models were exposed, shows that even sophisticated AI companies are targets, and that the consequences of a breach extend beyond the usual stolen credentials to the models themselves.

Managed Agentic Security

One pattern visible in the recent financings is the rise of managed, agentic security — security systems that use AI agents to monitor, detect and respond to threats, rather than relying solely on human analysts and static rules. Daylight Security, with its $80 million Series B, sits in this category, as do the other newly funded companies in different niches of the space.

This is not a replacement for human expertise but a force multiplier for it. The best security operations will combine agentic speed with human judgement on the hardest calls.

Why Incumbents and Startups Both Win

AI security is unusual in that it offers room for both established security vendors and new entrants. Incumbents have distribution, trust and existing customer relationships; they can add AI capabilities to products their customers already use. Startups have focus, speed and the freedom to design for the new threat landscape from scratch rather than retrofitting legacy architecture.

The venture rounds suggest investors believe startups have a genuine opening. An $80 million Series B and a $50 million Series A are not seed bets; they are growth-stage commitments to companies that have found customers and are scaling. At the same time, the largest security incumbents are investing heavily in their own AI capabilities, and the eventual market will likely be shared.

For founders, the question is where to position. Competing head-on with incumbents on their home turf is hard. Building for the threats incumbents are slow to address — agent security, model protection, AI supply-chain risk — offers a clearer path.

Diligence Questions for Founders in This Space

Founders building in AI security should expect tough questions, and should be able to answer them. Who are your design partners, and are they real users or friendly logos? Can you demonstrate that your system catches attacks that existing tools miss, rather than simply relabelling old alerts? How do you handle false positives, which are the bane of every security team? And how do you protect your own models and data, given that a security vendor that gets breached — as Fireworks AI's experience shows — faces a crisis of trust?

The Lumbra example is instructive here. Customers including Pentagon and intelligence community networks are the most demanding buyers in the world. Winning them is a strong signal, but keeping them requires a level of rigour that most startups have never faced.

The Bottom Line

AI security is hot because the need is real, the threat is growing, and the old tools are not enough. The money flowing into Daylight, Bay and Lumbra, set against the Fireworks AI breach, tells founders everything they need to know: build security into AI from the start, prove it works against real attacks, and the capital is there.

Angeled.com connects founders with investors.

Daylight Security Series B announcement ($80 million; Goldman Growth Equity, Bain, Assaf Rappaport)
Bay Series A announcement ($50 million, Battery; $13 million seed)
Lumbra Series A announcement ($25 million; Point72, Ansa)
Fireworks AI cyberattack disclosure, November 10, 2026

Read Next