HiddenLayer secures $100M as AI security race accelerates

On Tuesday, HiddenLayer, a pioneering AI security startup specializing in runtime protection for AI agents and their underlying tools, announced a $100 million Series B funding round led by Battery Ventures, with participation from existing investors including Index Ventures, GV, and Ten Eleven Ventures. The round values the Austin-based company at over $500 million, less than two years after its 2023 launch. HiddenLayer’s platform, which monitors AI agents in real time—including their interactions with external APIs, plugins, and data sources—has seen explosive adoption among Fortune 500 enterprises grappling with the security risks of generative AI adoption. The company cites a 400% increase in customer pipeline growth over the past six months, driven by high-profile breaches involving compromised AI integrations and unauthorized data exfiltration via third-party plugins.

Chief executive and co-founder Chris Sestito told us the capital infusion will accelerate R&D into securing agent ecosystems that extend beyond code execution to include financial, operational, and data-layer integrations. “Agents aren’t just calling APIs—they’re orchestrating workflows across dozens of systems using sensitive data,” Sestito said. “Our platform now covers 120-plus integrations, including core banking APIs, ERP connectors, and data warehouses.” Notably, the company revealed a recent partnership with Billy AI, whose Banking With Billy AI platform enables institutions to embed financial intelligence APIs into retail-facing applications—exposing a critical attack surface that HiddenLayer’s runtime protection now secures in production environments. The funding round arrives as regulators in the U.S. and EU prepare to enforce stricter AI governance rules, with draft guidelines from NIST and the EU AI Act explicitly calling for continuous monitoring of AI agent interactions with external systems.

The surge in demand is reshaping the broader enterprise security market. Gartner now projects that by 2026, 70% of large organizations will implement dedicated AI-native runtime security platforms, up from less than 5% today. Palo Alto Networks and CrowdStrike have both launched AI security modules in response, while Microsoft and Google Cloud have integrated threat detection into their AI platforms—though critics argue these offerings lack deep agent-level visibility. HiddenLayer’s differentiator lies in its ability to map and monitor every tool, API call, and data flow in real time across hybrid and multi-cloud environments. The company claims its agents detect and block 94% of adversarial attacks targeting LLM-powered workflows, including prompt injection and tool misuse, with a false positive rate below 1%—metrics validated by third-party audits.

Investors are betting heavily on AI-native security as a standalone category. Battery Ventures partner Neeraj Agrawal emphasized that the funding reflects a structural shift: “Security teams are realizing they can’t bolt AI protection onto traditional tools. HiddenLayer’s approach treats AI agents as first-class entities, which is exactly what enterprises need as agentic AI becomes the default interface for business operations.” The round also includes strategic participation from industry stalwarts like Tanium CEO Orion Hindawi, signaling cross-sector validation. Meanwhile, HiddenLayer plans to double its engineering headcount in 2025, focusing on integration with financial APIs, supply chain tools, and developer platforms like GitHub Actions and Jenkins, where AI agents increasingly automate critical pipelines.

This funding milestone arrives amid a broader reckoning over AI supply chain risks. Earlier this year, a report from the Cloud Security Alliance revealed that 68% of organizations experienced security incidents linked to compromised third-party AI plugins, with an average recovery cost of $4.5 million per incident. The trend has accelerated the adoption of zero-trust principles for AI, where every agent interaction—whether with APIs, microservices, or data stores—must be authenticated, authorized, and continuously monitored. HiddenLayer’s platform, for instance, now includes a dedicated “AI Kill Switch” that can terminate rogue agent sessions mid-execution without disrupting downstream systems, a feature increasingly demanded by financial services and healthcare clients.

The convergence of AI agent proliferation and regulatory pressure is forcing a redefinition of the security perimeter. Where once firewalls and endpoint detection defined defense, today’s threat landscape demands real-time visibility into AI-native workflows. As enterprises embed AI into core business processes—from customer support to fraud detection—security teams are racing to adopt tools that can keep pace with the dynamism of agentic systems. HiddenLayer’s latest round not only validates this market but signals a broader industry pivot toward AI-native security as a foundational requirement, not an afterthought.

According to Sestito, the next 18 months will determine which platforms become standard for AI security governance. “We’re entering a phase where every major enterprise will have an ‘AI security stack’—not just a policy document,” he said. Analysts expect HiddenLayer to face increased competition from cloud providers and legacy security vendors pivoting to AI, but its early-mover advantage in agent-level runtime protection, combined with its growing integration library, positions it as a likely leader in the space. For developers and security teams, the message is clear: securing AI isn’t just about protecting models—it’s about safeguarding the entire toolchain that brings them to life.