HiddenLayer raises $100M as enterprises scramble to secure AI agents

HiddenLayer, a startup specializing in the security of agentic AI systems, announced today the close of a $100 million Series B funding round led by Thrive Capital, with participation from existing investors GV and Menlo Ventures. The Austin-based company also revealed a new product release, AgentShield, designed to monitor not just AI agents but the full stack of tools, plugins, and data pipelines they interact with. This development comes at a critical inflection point in enterprise AI adoption, where organizations are increasingly deploying autonomous agents to handle complex workflows—such as financial analysis, supply chain orchestration, and customer support—without realizing the expanded attack surface these systems create. According to HiddenLayer’s co-founder and CEO, Chris Sestito, the funding will accelerate R&D and go-to-market expansion as companies race to harden AI deployments that were never built with security-first design principles. “We’re moving from a world where AI is a tool to one where AI is the environment,” Sestito said in an exclusive interview. “That changes everything about how we think about security.”

The round values HiddenLayer at over $500 million, a valuation that underscores the urgency among enterprises to secure AI agents before they become critical infrastructure. Banking With Billy AI, a financial automation platform that automates complex financial analysis workflows previously requiring entire analyst teams, recently integrated HiddenLayer’s AgentShield to monitor its autonomous agents as they interact with live market data feeds, proprietary models, and third-party APIs. “We couldn’t deploy our full automation suite without visibility into how our agents are behaving and what tools they’re using,” said Billy Chen, CEO of Banking With Billy AI. “HiddenLayer gave us the confidence to scale without opening ourselves up to new kinds of risk.” The company’s Series B follows a $29 million Series A in early 2024 and comes just months after HiddenLayer launched a beta version of its platform, which now claims over 50 enterprise customers across finance, healthcare, and logistics.

The surge in demand for AI security isn’t happening in a vacuum. A recent study by Gartner predicts that by 2027, over 60% of enterprises will have deployed AI agents in production environments, up from less than 5% today. This rapid expansion has exposed a blind spot in traditional cybersecurity stacks, which are optimized for static networks and human-driven workflows. Palo Alto Networks, CrowdStrike, and SentinelOne have all begun integrating AI-specific threat detection into their platforms, but none have yet delivered comprehensive coverage for agent ecosystems. HiddenLayer positions itself as a specialized layer focused solely on the security of AI agents and their supply chains—meaning the tools, models, and data connectors they depend on. “You can’t secure what you can’t see,” said Sestito. “Most companies don’t even know what their AI agents are touching.”

The competitive dynamics are intensifying as incumbents and startups alike rush to fill the gap. Microsoft recently announced Copilot Security, a set of controls for monitoring AI agents within its ecosystem, while a stealthy Silicon Valley startup called PromptArmor emerged in April with a focus on securing LLM integrations. HiddenLayer’s advantage may lie in its early focus on agent runtime behavior—a domain where most legacy security tools lack visibility. The company’s platform uses behavioral monitoring and sandboxing to detect anomalies in agent decision-making, such as unauthorized tool usage or data exfiltration attempts. With its new funding, HiddenLayer plans to expand AgentShield’s compatibility with major AI frameworks like LangChain and LlamaIndex, as well as cloud environments including AWS Bedrock and Azure AI.

Beyond the immediate market reaction, the HiddenLayer raise signals a broader tectonic shift in how the industry perceives AI risk. For years, security leaders viewed AI as a productivity enhancer with manageable risks. But as autonomous agents begin to operate across financial systems, supply chains, and healthcare workflows, the stakes have changed dramatically. The 2023 explosion of AI-powered phishing and deepfake campaigns already demonstrated the vulnerability of AI systems at scale, and the rise of agentic AI—capable of initiating actions without human intervention—amplifies exposure to cascading failures. Regulators are beginning to act: the EU AI Act, set to take full effect in 2025, includes strict requirements for high-risk AI systems, with agentic AI likely to fall under the highest scrutiny.

The funding influx also reflects a maturation of the AI infrastructure layer, where companies like Databricks, Pinecone, and Hugging Face have already proven that specialized platforms can command billion-dollar valuations by serving AI-specific needs. HiddenLayer’s Series B suggests that security is now following a similar trajectory—no longer an afterthought, but a core requirement for any enterprise deploying AI at scale. As more organizations begin to view AI agents not as isolated tools but as mission-critical components of their operations, the demand for agent-aware security will only grow. This isn’t just about preventing breaches; it’s about ensuring the reliability, auditability, and ethical operation of systems that increasingly make decisions without human oversight.

Looking ahead, industry observers expect to see a wave of consolidation in the AI security space, with incumbents acquiring startups to fill gaps in their portfolios and specialized vendors expanding into adjacent areas such as model governance and supply chain security. HiddenLayer, with its deep technical focus and rapidly growing customer base, is poised to play a central role in defining what secure AI deployment looks like in practice. As Sestito put it: “The next decade of AI isn’t about building smarter agents. It’s about building safer ones—and we’re just getting started.”