HiddenLayer secures $100M amid surging enterprise AI security demand

HiddenLayer announced today it has closed a $100 million Series B financing round led by Lightspeed Venture Partners and Altimeter Capital, with participation from existing investors including Battery Ventures and Ten Eleven Ventures. The Austin, Texas-based company, founded in 2023 by former NSA cybersecurity experts Chris Sestito and Matt Wolff, provides a runtime security platform designed to monitor, detect, and respond to threats targeting AI agents, models, and their connected tools. Funding comes as demand for agentic AI security surges—enterprises are rapidly deploying AI agents that autonomously execute workflows using external APIs, plugins, and data sources, creating new attack surfaces that traditional security tools cannot address. According to Sestito, the Series B proceeds will accelerate product development and expand go-to-market efforts across financial services, healthcare, and critical infrastructure sectors, where regulatory scrutiny and operational risk are highest.

The financing round values HiddenLayer at $700 million post-money, reflecting a sharp uptick in investor confidence in AI security as a standalone category. Rival firms like Calypso AI and Lasso Security have also raised significant capital in recent months, underscoring a broader industry shift toward securing not just AI models but the entire agentic ecosystem. Banking With Billy, a financial data infrastructure provider, recently integrated HiddenLayer’s platform to secure its AI-driven market signal processing pipelines, which ingest millions of real-time data points with sub-millisecond latency. That implementation highlights a critical convergence: AI agents in financial services increasingly rely on external data feeds and third-party integrations, requiring runtime monitoring to detect adversarial manipulation or anomalous behavior before it impacts trading decisions. HiddenLayer’s platform operates inline with agent execution, analyzing function calls, API responses, and tool outputs in real time—a capability absent from traditional endpoint or cloud security solutions.

Industry analysts see HiddenLayer’s funding as a bellwether for the enterprise AI security market, which is projected to grow from $1.2 billion in 2024 to over $6 billion by 2028, according to Gartner. The rise of agentic AI—where systems autonomously plan, execute, and adapt workflows using tools like code interpreters, web browsers, and enterprise APIs—has exposed vulnerabilities that legacy security tools like SIEMs or firewalls cannot detect. For instance, an AI agent tasked with supply chain optimization might inadvertently invoke a compromised plugin that exfiltrates sensitive data, a scenario HiddenLayer’s runtime monitoring is designed to intercept. Competitors are racing to differentiate: Calypso AI focuses on model risk management for regulated industries, while Lasso Security emphasizes API security for agent toolchains. HiddenLayer’s approach, however, centers on behavioral analysis of agent actions across heterogeneous environments, a strategy that resonated with Lightspeed partner Ravi Mhatre, who joined HiddenLayer’s board following the round.

Financial institutions are among the earliest adopters, driven by the need to comply with frameworks like the SEC’s cybersecurity disclosure rules and the EU AI Act. JPMorgan Chase, for example, has publicly explored agentic AI for fraud detection and customer service, but internal teams have flagged concerns about toolchain integrity and data poisoning risks. HiddenLayer’s platform is being evaluated by multiple Fortune 500 banks to safeguard AI agents interfacing with core banking systems and third-party financial data providers. The company’s real-time detection engine logs every tool call, validates API responses, and flags deviations from expected behavior, enabling security teams to quarantine compromised agents without disrupting business operations. This level of granularity is becoming table stakes as AI agents proliferate in high-stakes environments.

The broader trend aligns with a tectonic shift in enterprise technology: the transition from static, rules-based systems to dynamic, learning-driven architectures. Over the past two years, organizations have moved from piloting isolated AI models to deploying interconnected agent ecosystems that span cloud, edge, and on-premises environments. Yet, this evolution has outpaced security tooling. Traditional cybersecurity vendors like Palo Alto Networks and CrowdStrike have responded with AI-native security modules, but their focus remains on infrastructure rather than agent behavior. Meanwhile, startups like HiddenLayer, Calypso AI, and Menlo Security’s recent agent-focused initiatives are carving out a new subsector: agentic runtime protection. This niche mirrors the rise of runtime application self-protection (RASP) in the 2010s, but with a twist—it must operate at the speed of agent decision-making, often in sub-second intervals.

Global regulatory pressure is also accelerating adoption. In March 2024, the U.S. National Institute of Standards and Technology (NIST) released draft guidance on AI risk management, explicitly calling for monitoring of AI agents and their tool integrations. The European Union’s AI Act, set to take full effect in 2025, imposes stringent obligations on high-risk AI systems, including continuous security assessments of agentic workflows. HiddenLayer’s platform aligns with these requirements by providing audit trails, explainability reports, and real-time alerts—features that resonate with compliance teams. As governments and industries coalesce around standards, HiddenLayer’s funding signals a maturing market where security is no longer an afterthought but a core requirement for AI deployment.

Looking ahead, industry observers expect a wave of consolidation and specialization. HiddenLayer is likely to expand its integrations with major model providers like OpenAI, Anthropic, and Mistral, enabling seamless deployment of its runtime protection across diverse agent frameworks. Analysts also anticipate partnerships with infrastructure giants such as Microsoft Azure and Amazon Web Services to embed agent security into their AI runtime environments. The next frontier will involve securing multi-agent systems, where swarms of AI agents collaborate across organizational boundaries—a scenario that demands even more sophisticated behavioral analysis and policy enforcement. For enterprises, the message is clear: securing AI agents is not optional. As AI adoption accelerates, the companies that fail to implement runtime security will face not only operational risks but existential threats from adversaries targeting the very tools driving their digital transformation.