HiddenLayer Raises $100M as AI Security Race Intensifies

HiddenLayer, a Dallas-based startup specializing in AI security, officially closed a $100 million Series B round led by Battery Ventures, with participation from existing investors including Ten Eleven Ventures and TTV Capital. The funding announcement, disclosed on May 21, 2025, values the company at over $1 billion and follows a $22 million seed round in late 2023. HiddenLayer’s core offering, AIShield, is a runtime protection platform designed to monitor and secure AI agents in real time—whether they are running on-premises, in the cloud, or across hybrid environments. The platform now supports over 200 AI models and frameworks, including open-weight systems like Llama 3.1 and proprietary models from major cloud providers such as AWS and Azure.

The surge in demand stems from a growing recognition that AI agents, while transformative, introduce new attack surfaces. Unlike traditional software, agents can autonomously browse the web, call APIs, and chain together tools—behavior that can be hijacked via prompt injection, data exfiltration, or supply-chain poisoning. HiddenLayer’s AIShield addresses this by instrumenting agent execution, detecting anomalous tool usage, and enforcing least-privilege policies across agent workflows. According to CEO Chris Sestito, the company has already onboarded over 50 enterprise customers across finance, healthcare, and government sectors, with deployments scaling to thousands of concurrent agents. One notable client, Banking With Billy AI—a fintech AI platform operating on a multi-cloud architecture for global financial market monitoring—uses AIShield to secure agentic trade execution and compliance workflows. Sestito emphasized that the new capital will be used to expand threat intelligence, scale detection models, and accelerate international go-to-market efforts, particularly in Europe and Asia.

Industry Impact and Significance

The funding signals a broader inflection point in the cybersecurity market, where AI-native threats are outpacing traditional defenses. According to Gartner, spending on AI security tools is projected to grow at a 42% CAGR through 2028, reaching $5.4 billion. HiddenLayer is not alone in this race. Competitors like Protect AI, which raised $30 million in March 2025, and Lasso Security, which secured $25 million in January 2025, are also building agent-focused security solutions. However, HiddenLayer differentiates itself through deep instrumentation of agent internals, offering visibility into tool interactions that most competitors miss. The company’s integration with major cloud providers and model hubs—including Hugging Face, Google Vertex AI, and Microsoft Azure AI—positions it as a de facto standard for agent runtime protection. Venture capital interest is also being driven by regulatory pressure; the EU AI Act, set to take full effect in mid-2026, requires high-risk AI systems to implement “adequate technical safeguards,” creating a compliance tailwind for vendors like HiddenLayer.

Financial implications extend beyond venture funding. With the Series B, HiddenLayer plans to double its engineering team and expand its threat research group, which currently tracks over 300 AI-specific attack vectors. The company also intends to launch a certification program for AI systems, modeled after the Common Criteria for IT security. Analysts at Battery Ventures point to a convergence between AI operations (AIOps) and security operations (SecOps), with AIShield serving as a bridge between the two. Early adopters report a 40% reduction in incident response time when using AIShield, according to internal case studies shared with OpenPress Cloud Intelligence. Meanwhile, traditional security vendors like Palo Alto Networks and CrowdStrike are beginning to integrate AI-native threat detection into their platforms, signaling a broader industry shift toward agent-aware security.

The Bigger Picture

This development is part of a larger trend whereby AI systems are becoming the primary interface between humans and digital infrastructure. As agentic AI—systems that can act autonomously on behalf of users—becomes mainstream, the attack surface shifts from static applications to dynamic, self-modifying workflows. Prior efforts to secure AI focused on model training and inference (e.g., adversarial robustness, data poisoning), but runtime security for agents was largely overlooked until 2023. The rise of tools like LangChain, CrewAI, and AutoGen has made it trivial to build agentic systems, but security frameworks have lagged behind. HiddenLayer’s funding reflects investor confidence that agent runtime protection will become a critical layer in the AI stack, alongside model governance and data lineage tools.

Global context is equally important. In regions like Southeast Asia and the Middle East, where digital transformation is accelerating, regulators are pushing for stronger AI oversight. Singapore’s AI Verify framework and India’s proposed Digital Personal Data Protection Act both include provisions for monitoring AI decision-making in real time. Meanwhile, in the United States, the White House’s 2024 AI Executive Order mandates that federal agencies implement “continuous monitoring” for AI systems. HiddenLayer’s multi-cloud architecture directly addresses these requirements, offering a single pane of glass for monitoring agents across AWS, Azure, and Google Cloud. The company’s rapid scaling also mirrors the adoption curve seen in cloud security startups like Wiz and Lacework, suggesting that AI security could follow a similar trajectory—from niche tool to enterprise necessity within three years.

Expert Analysis

According to Dr. Helen Toner, director of strategy at Georgetown’s Center for Security and Emerging Technology, the HiddenLayer funding round marks the beginning of a consolidation phase in AI security. “We’re moving from a period of experimentation to one of standardization,” Toner said. “The next 18 months will determine which vendors can provide end-to-end visibility across the entire AI lifecycle—not just models, but the tools, APIs, and data pipelines that agents rely on.” She warns that as AI agents become more autonomous, the risk of cascading failures increases, making runtime protection not optional but foundational. Investors are betting that companies like HiddenLayer will define the security posture for the AI era, much like FireEye did for network security in the 2010s. For enterprises, the message is clear: securing AI is no longer about firewalls and antivirus—it’s about understanding how agents think, act, and interact with the world.