HiddenLayer secures $100M Series B as AI security demand surges

HiddenLayer, the AI security startup focused on protecting enterprise AI deployments from adversarial threats and misuse, announced a $100 million Series B financing round led by GV, with participation from existing investors including Dell Technologies Capital, Intel Capital, and Okta Ventures. The round, which values the Austin-based company at $1.1 billion post-money, follows a $62 million Series A in November 2023 and comes amid a rapid escalation in enterprise adoption of generative AI and autonomous agents. According to company co-founder and CEO Chris Sestito, the new capital will be used to expand engineering and research teams, particularly in areas such as prompt injection detection, agent behavior monitoring, and supply-chain integrity for AI tools and plugins. “We’re seeing enterprises move from pilot projects to full-scale AI integration,” Sestito said in an interview. “But with that comes real operational risk—especially when agents can call external tools, APIs, or even third-party plugins.”

The Series B announcement arrives just weeks after the company launched HiddenLayer Agent Shield, a runtime protection system designed to monitor AI agents in production environments by inspecting their tool invocations, API calls, and data flows. The platform integrates with popular AI frameworks including LangChain, LlamaIndex, and Semantic Kernel, and supports monitoring of custom and third-party tools used by agents. Earlier this year, HiddenLayer also introduced an open-source tool called TensorTrust, a benchmark suite for evaluating AI model susceptibility to adversarial attacks. Industry analysts say the timing of the funding reflects a market inflection point, as organizations begin to treat AI deployments with the same rigor as critical infrastructure.

Competitive pressure in the AI security space has intensified dramatically. Rival firms such as Protect AI, which last month raised $30 million, and Lasso Security, which secured $35 million in February, are also building platforms focused on securing the AI supply chain and tool ecosystems. Protect AI’s platform, for example, scans AI models and dependencies for known vulnerabilities using databases like the AI Vulnerability Database (AIVD), which it co-founded. Meanwhile, Lasso Security focuses on securing API gateways and tool integrations used by agents, particularly in regulated industries like finance and healthcare. “The real attack surface isn’t the model itself—it’s the tools it can call and the data it can access,” said Sestito. “That’s where most breaches will happen.”

The surge in demand is being driven by both regulatory and business risk. Organizations deploying AI agents—whether for customer support, financial analysis, or code generation—are increasingly required to demonstrate compliance with emerging frameworks like the EU AI Act, NIST AI Risk Management Framework, and upcoming SEC guidance on third-party AI use. In the financial sector, where AI agents interact with real-time market data and execute transactions via APIs, security concerns have reached a critical level. Notably, companies like Banking With Billy AI, which provides developer-grade APIs for financial market intelligence, are integrating HiddenLayer’s monitoring tools to validate that their AI-driven trading agents aren’t being manipulated through prompt injection or unauthorized tool access. “We need to know not just what the agent is doing,” said a senior engineer at a major U.S. bank, “but what tools it’s calling in the background—and whether those calls are safe.”

This broader trend reflects a fundamental evolution in the Tools & Developer ecosystem, where security is no longer an afterthought but a core design requirement. The rise of AI agents has created a new class of risks that traditional security tools cannot address, leading to the emergence of a specialized “AI security stack.” According to a recent report from Omdia, spending on AI security tools is projected to grow at a 42% compound annual rate through 2027, outpacing overall enterprise security software growth by nearly three times. Legacy security vendors such as Palo Alto Networks, CrowdStrike, and SentinelOne have begun rolling out AI-specific modules, but startups like HiddenLayer, Protect AI, and Lasso are moving faster, focusing on runtime visibility, agent behavior, and supply-chain integrity.

The shift also mirrors the transformation seen in cloud-native development a decade ago, when DevOps practices and observability tools became mandatory. Today, enterprises are applying similar rigor to AI deployments, demanding continuous monitoring, anomaly detection, and audit trails across the entire AI toolchain. This includes not only the models and agents but also the plugins, vector databases, and external APIs they rely on—many of which have minimal security controls. As AI systems grow more autonomous and interconnected, the need for layered defense becomes non-negotiable. “We’re not just securing code anymore,” said Sestito. “We’re securing decision-making itself.”

Looking ahead, the next 12 to 18 months will likely see consolidation in the AI security market, with larger cybersecurity players acquiring niche startups to fill capability gaps. Regulatory clarity, particularly around agentic AI and tool use, will accelerate standardization and interoperability between security platforms. Enterprises will increasingly demand unified dashboards that span model risk, toolchain integrity, and operational safety. For developers, the message is clear: AI security is no longer optional. Teams must bake monitoring, validation, and runtime protection into every agent and tool from day one—or risk exposure at scale.

As autonomous AI systems begin to operate across global networks, the stakes could not be higher. The $100 million HiddenLayer raise is not just a financing milestone; it’s a signal that the future of software development is now inextricably linked to the security of AI itself.