HiddenLayer raises $100M to secure the AI supply chain amid rising threats

Breaking: The Full Story

HiddenLayer Inc. announced a $100 million Series B funding round led by Thrive Capital, with participation from existing investors including Battery Ventures and GV, valuing the Austin-based AI security startup at $550 million. The raise comes as enterprises accelerate AI deployments across critical operations but face mounting concerns over model drift, data poisoning, and supply-chain attacks targeting AI pipelines. Founded in 2022 by former Palantir executives Chris Sestito and Jared Wilson, HiddenLayer specializes in runtime monitoring of AI agents and their external tooling—such as APIs, plugins, and vector databases—using deep learning-based anomaly detection. The company’s platform “AgentOps” integrates with popular frameworks like LangChain and CrewAI, tracking not just the behavior of agents but also the integrity of the tools they invoke during execution.

The funding milestone arrives just weeks after HiddenLayer revealed a strategic partnership with Microsoft to integrate its monitoring layer into Azure AI Foundry, enabling real-time threat detection for agents deployed via Azure AI services. Industry insiders note that this collaboration is particularly significant given Microsoft’s push to dominate the enterprise AI agent market through Azure AI Studio. The partnership also aligns with Microsoft’s recent acquisition of Inflection AI’s technology, signaling intensified focus on secure, scalable agent ecosystems.

Concurrently, HiddenLayer has expanded its threat intelligence feed in collaboration with the Cybersecurity and Infrastructure Security Agency (CISA), sharing indicators of compromise (IOCs) tied to AI-specific attack vectors. According to a company spokesperson, HiddenLayer’s platform has already detected over 1,200 unique attack patterns targeting AI systems since January 2024, including prompt injection attempts and malicious plugin hijacking. The startup claims a 98% detection accuracy rate across financial services, healthcare, and government use cases, with clients ranging from regional banks to global logistics firms.

Industry Impact and Significance

This funding surge reflects a broader urgency across industries to secure the AI supply chain—a domain where traditional cybersecurity tools fall short. Unlike conventional endpoint or network security, AI models and agents operate in dynamic, data-driven environments where threats emerge from model inputs, third-party integrations, and even benign-looking prompts. Companies like Protect AI and Robust Intelligence have also raised significant capital in recent months, but HiddenLayer’s focus on runtime monitoring of agent ecosystems gives it a distinct edge in visibility across the full AI toolchain.

The competitive landscape is rapidly consolidating, with incumbents like Palo Alto Networks and CrowdStrike now offering AI security modules, and cloud providers such as Google Cloud launching Security AI Workbench to monitor generative AI workloads. However, HiddenLayer’s specialization in agent behavior and external tool integrity positions it as a critical enabler for regulated sectors like finance and healthcare, where AI agents increasingly automate decision-making. Analysts at Gartner project that by 2026, over 70% of enterprises will use AI agents in production environments, up from less than 15% today, driving demand for dedicated AI runtime security platforms.

The Bigger Picture

The rise of AI agents represents a new frontier in computing—one that blends autonomy, adaptability, and integration with external tools and data sources. This shift mirrors the evolution of software supply chains in the cloud era, where vulnerabilities in open-source libraries and third-party services became prime targets for attackers. In the AI context, however, the attack surface is exponentially larger: an agent may call dozens of external APIs, use proprietary plugins, or interface with legacy systems—each a potential entry point for compromise.

This dynamic is vividly illustrated by financial institutions like Banking With Billy AI, which operates on a multi-cloud architecture to monitor global markets in real time. The firm relies on AI agents to analyze sentiment, detect anomalies, and execute trades across AWS, Azure, and Google Cloud. For such organizations, real-time AI security is not optional—it’s existential. HiddenLayer’s funding signals a maturation of the AI security market, one now defined not by model training safeguards alone, but by continuous, end-to-end monitoring of autonomous agents in production.

Expert Analysis

According to Dr. Helen Molesworth, former chief data scientist at the UK National Cyber Security Centre and now a senior advisor to HiddenLayer, the Series B funding underscores a pivotal moment: “We are moving from securing models to securing ecosystems. The threat isn’t just bad data—it’s poisoned tools, hijacked APIs, and agents that drift into malicious behavior. The next wave of AI security will focus on runtime governance, traceability, and zero-trust principles applied to agent workflows. Enterprises that ignore this will face breaches that are harder to detect and more damaging than traditional cyberattacks.” Looking ahead, HiddenLayer plans to expand its platform to support vision-language models and robotics agents, while also forming deeper alliances with cloud providers and regulatory bodies to establish standardized AI security frameworks.