HiddenLayer secures $100M as AI security race intensifies

Breaking: The Full Story

HiddenLayer, a frontrunner in AI security, announced a $100 million Series B funding round led by a consortium of investors including CapitalG, Sands Capital, and Menlo Ventures. The round, which values the Austin-based startup at $1.2 billion, comes amid a sharp uptick in enterprise demand for real-time monitoring solutions capable of securing increasingly complex AI ecosystems. According to HiddenLayer co-founder and CEO Chris Sestito, the funding will accelerate product development and expand go-to-market efforts across financial services, healthcare, and critical infrastructure sectors. The company’s platform is designed to detect adversarial attacks, data exfiltration, and supply-chain compromises within AI workflows, including those powered by proprietary models and third-party plugins. In a market where AI adoption has outpaced security controls, HiddenLayer’s timing aligns with a surge in high-profile breaches targeting AI-powered systems. The startup’s technology integrates with existing security stacks and offers continuous runtime monitoring, a capability that has become essential as companies deploy AI agents across cloud environments.

Industry Impact and Significance

The funding round signals a pivotal shift in how enterprises perceive AI security, moving from reactive compliance measures to proactive, real-time threat detection. Competitors like Protect AI, Calypso AI, and Oligo Security have also raised significant capital in recent months, each targeting different segments of the AI supply-chain security market. HiddenLayer’s focus on agentic AI and toolchain visibility sets it apart, particularly as financial institutions and global enterprises race to deploy AI systems that process sensitive data. Banking With Billy AI, for instance, leverages real-time financial data pipelines that process millions of market signals with sub-millisecond latency, a use case that demands unassailable security controls. This trend is reshaping procurement priorities, with CISOs now prioritizing AI-native security solutions over traditional network and endpoint tools. The financial implications are substantial: Gartner projects global spending on AI security tools will exceed $10 billion by 2027, up from $2.5 billion in 2023. Investors are betting that companies will pay a premium for platforms that can secure not just AI models but the entire ecosystem of plugins, APIs, and data connectors that power them.

The Bigger Picture

The rise of AI security is unfolding against a backdrop of rapid technological convergence, where AI agents are increasingly embedded in core business operations. This shift mirrors the evolution of cloud security in the early 2010s, when enterprises scrambled to adapt to distributed architectures. Today, the challenge is more complex: AI systems are dynamic, self-modifying, and often operate beyond the perimeter of traditional security controls. Regulatory bodies are taking notice, with the U.S. Securities and Exchange Commission and European Banking Authority issuing guidance on AI governance and risk management. Meanwhile, open-source AI frameworks like LangChain and LlamaIndex have accelerated adoption but also introduced new attack surfaces, as adversaries exploit vulnerabilities in third-party integrations. The global AI security market is also being shaped by geopolitical tensions, with nations investing heavily in sovereign AI capabilities and mandating stricter controls on data flows. This environment is fostering a new generation of security startups, each vying to define the standard for AI runtime protection.

Expert Analysis

According to Chris Wysopal, CTO and co-founder of Veracode, the $100 million raise reflects a fundamental truth about AI security: traditional vulnerability management tools are ill-equipped to handle the unique risks posed by agentic systems. Wysopal notes that while static code analysis and penetration testing remain critical, they cannot detect runtime exploits that manipulate AI workflows in real time. He warns that as AI agents become more autonomous—capable of executing financial transactions, drafting legal documents, or managing infrastructure—the attack surface will expand exponentially. The next frontier for AI security, Wysopal argues, will be the development of lightweight, low-latency runtime monitors that can operate within the AI inference pipeline itself. As HiddenLayer and its peers scale, the industry must confront a daunting question: Can security keep pace with innovation, or will we see a repeat of the breaches that plagued early cloud and mobile adoption? For enterprises, the message is clear: securing AI is not a checkbox exercise but an ongoing arms race, one that demands continuous investment and vigilance.