AI security startup HiddenLayer has raised a $100 million Series B round to expand its platform for securing enterprise AI deployments, with a particular focus on AI agents that operate autonomously in production environments.

The round was led by Delta-v Capital, with participation from Ten Eleven Ventures, Morgan Stanley, M12 (Microsoft’s venture fund), and Booz Allen Ventures. It brings HiddenLayer’s total funding to approximately $156 million.

The Problem: AI Agents Are Now a Security Attack Surface

For the past few years, enterprise AI security mostly meant protecting the models themselves from adversarial inputs and data poisoning. That problem hasn’t gone away. But a new category has emerged: what happens after you deploy an AI agent and it starts taking actions on your behalf?

Autonomous AI agents now read emails, write and execute code, query databases, call APIs, and in some cases deploy software to production systems. That’s a lot of surface area for things to go wrong. An agent can be manipulated by malicious input, misuse tools in ways that weren’t intended, or take actions that exceed its sanctioned scope.

A recent survey found that 65% of enterprises have already seen AI agents act out of scope. That figure will only grow as agent deployments scale.

HiddenLayer’s core product, Agentic Runtime Security, gives organizations visibility into how their AI agents behave once they’re running in production. It flags and stops manipulation, tool misuse, and unauthorized actions as they happen rather than detecting problems after the damage is done.

At the Series B, the company also announced a new product specifically designed to secure AI coding agents. Tools like Cursor, GitHub Copilot, and Claude Code now write and ship code largely autonomously in many organizations. Each of those agents has read access to codebases, write access to files, and sometimes the ability to execute commands and push to repositories. A compromised or misbehaving coding agent could introduce vulnerabilities at scale, faster than any human reviewer could catch.

Why This Funding Round Matters

The investors here tell a story about where enterprise AI security is heading. Microsoft’s M12 putting money into an AI agent security platform is notable given that Microsoft is simultaneously one of the largest AI infrastructure providers. Booz Allen Ventures signals serious interest from the defense and government contracting sector, where AI agents are being deployed into high-stakes environments.

Delta-v Capital leading the round is consistent with their thesis around the security infrastructure needed to support the next generation of enterprise software.

The $100 million figure also reflects the market opportunity. Enterprise AI deployments have moved from pilots to production at speed over the past 18 months. Companies that rushed to capture productivity gains from AI agents are now discovering they need the governance infrastructure to go with them.

What This Means for Business

If your organization is running AI agents in any capacity, whether for customer service, data processing, coding assistance, or operations, this funding round is a signal worth paying attention to.

The gap between deploying AI agents and having runtime visibility into what those agents actually do is real, and it’s being taken seriously by serious investors.

A few practical considerations:

You probably don’t have runtime monitoring yet. Most organizations that adopted AI agents in 2024 and 2025 focused on getting them working, not on monitoring what they do once they’re live. That oversight gap is where HiddenLayer and competitors play.

Coding agents are a specific new risk. If your development teams are using AI coding tools with any level of autonomy (automated test runs, code commits, infrastructure changes), those tools need the same scrutiny as any other system with write access to your infrastructure.

This becomes a vendor evaluation question. As the AI security category matures, expect it to appear on enterprise security questionnaires and procurement checklists. If you’re deploying AI agents for customers, having an answer to “how do you monitor agent behavior in production?” is no longer optional.

The regulatory environment is moving in the same direction. Multiple jurisdictions now have or are developing requirements around AI system monitoring, audit trails, and incident reporting. Runtime security infrastructure like HiddenLayer’s is the practical implementation of those requirements.

The speed of AI adoption in enterprise environments has consistently outpaced the governance and security infrastructure to support it. HiddenLayer is building that infrastructure. The fact that it has attracted $100 million in fresh capital suggests the market agrees that catching up is now urgent.

---

For businesses looking to deploy AI agents with proper governance frameworks, Omni Ops from Enterprise DNA includes implementation guidance on monitoring and responsible agent deployment as part of every engagement.