AI Agent Security: Neo Launches With $100M to Lock It Down
A new security startup named Neo left stealth on July 20 with $100 million in funding, the company announced, aiming to secure the flood of AI agents now entering the enterprise. Andreessen Horowitz and Bessemer Venture Partners led the round, with Craft Ventures and Merlin Ventures joining.
The size of the check signals where investor attention is heading. As companies wire AI agents into daily operations, each new agent becomes a fresh entry point for attackers. That is why AI agent security is moving from a niche worry to a board-level line item, a shift that also touches identity security.
Who Is Behind Neo
The founding team comes from three of the best-known names in the field. According to the company, its leaders previously worked at SentinelOne, Wiz, and Palo Alto Networks. That pedigree helps explain the nine-figure raise for a company most people had never heard of a week ago.
Experience matters in security, because buyers trust teams that have shipped at scale. Investors clearly bet that this group can move fast in a market that is expanding by the month.
Still, a strong resume is not a product. The real test is whether Neo can deliver before rivals crowd the same space.
What the Platform Does
Neo targets a simple but growing problem: security teams cannot protect what they cannot see. Its platform helps them find every AI agent, application, browser, and identity operating inside a company, then apply rules to each one.
From there, the tool tracks risk and traces actions back to their source. That combination of visibility and control is meant to give security staff a single place to manage the sprawl of AI software across the business.
| Detail | Figure |
|---|---|
| Funding raised | $100 million |
| Lead investors | Andreessen Horowitz, Bessemer |
| Launch date | July 20, 2026 |
The pitch is consolidation. Rather than bolting on a new tool for every AI system, teams get one place to watch agents, apps, and identities together. For a lean team, fewer blind spots often beats another point solution.
Why the Timing Works
The market backdrop explains the urgency. Gartner reports that fewer than 5 percent of enterprise applications carried agentic capabilities in 2025, but it expects that figure to reach 40 percent by the end of 2026. That is an eightfold jump in a single year.
Every one of those agents can read data, take actions, and connect to other systems. As a result, the attack surface grows with each rollout, and old security tools were not built for software that acts on its own.
The money follows that math. Investors are backing the companies that clean up after fast AI adoption, because every rushed rollout creates new gaps to close. So the founder-relevant takeaway is clear: if your product uses AI agents, security is now part of the roadmap, not an afterthought.
What Founders Should Watch
Most young companies will not buy an enterprise platform like Neo yet. Even so, the launch flags a risk that applies to teams of any size. Agents that touch customer data or money need guardrails before they scale.
Practical steps cost little. Limit what each agent can access, log its actions, and review permissions often. Treat every new agent like a new employee with keys to the building, and give it only the access it truly needs, which mirrors good AI governance.
The Bigger Signal for AI Builders
Neo is one data point in a broader pattern. Money is flowing to companies that clean up the messes created by fast AI adoption. Security, monitoring, and compliance are becoming their own markets.
For founders, that spells opportunity as much as caution. The same wave that raises the stakes also opens room for tools that make AI safer, a gap felt even in small business AI adoption. Trust, in other words, is becoming a product category of its own.
AI Agent Security FAQ
Why did Neo raise $100 million? The company plans to scale its platform as enterprises rush to adopt AI agents, which create new security gaps that older tools were not designed to cover.
Do small startups need AI agent security? Yes, at a basic level. Any agent that can access customer data or funds should have limited permissions, activity logs, and regular reviews, even on a small team.
What is agentic AI, and why is it a security risk? Agentic AI can act on its own, calling tools and moving data without a human in the loop, which widens the ways an attacker can get in.
