AI News · $100M+ rounds ·
HiddenLayer raises $100 million to expand AI security development
HiddenLayer raised $100 million in Series B funding led by Delta-v Capital. Participants included Ten Eleven Ventures, Morgan Stanley, Microsoft's M12 and Booz Allen Hamilton, as enterprises seek to secure their AI deployments.
Key points
- HiddenLayer raised $100 million in Series B funding led by Delta-v Capital.
- The funding will support sales, distribution, engineering, research and geographic expansion.
- Its products now address prompt injection, agent manipulation and malicious tool use.
- CEO Chris Sestito reported more than tenfold annual recurring revenue growth but did not disclose an exact figure.
- Some capabilities could eventually be bundled into larger AI platforms, Sestito acknowledged.
What happened: HiddenLayer has raised $100 million in Series B funding led by Delta-v Capital, TechCrunch reported. Participants included Ten Eleven Ventures, Morgan Stanley, Microsoft’s M12 and Booz Allen Hamilton. The Austin-based startup builds security tools to protect AI models, agents and workflows against adversarial attacks, vulnerabilities and malicious code injections. The company said the funding will support a greater focus on sales and distribution while it continues expanding engineering and research. It also plans to expand into Europe and the wider EMEA region.
The details: HiddenLayer’s products cover discovery, runtime protection, attack simulation and supply chain security. Co-founder and CEO Chris Sestito told TechCrunch that the company has extended those products to address prompt injection, agent manipulation and malicious tool use as customers move from traditional machine learning to generative AI and agents. He highlighted runtime security, which protects AI during operation, as an increasing priority and compared it with endpoint detection and response tools built specifically for AI. Sestito also said HiddenLayer scans about 50 AI file frameworks, looking for issues such as models that misrepresent what they are or contain hidden models.
Background: The financing comes three years after HiddenLayer’s $50 million Series A, when the scale of demand for dedicated AI security tools was less clear, according to TechCrunch. The outlet reported that security companies are now building products to monitor agents alongside the tools and add-ons they use. Gartner estimates that businesses will spend $2.83 billion on products to secure AI tools in 2026, up 83% from 2025, with spending expected to reach nearly $4.78 billion in 2027. Those figures are forecasts, rather than measured future spending, and TechCrunch noted that public reports of agents being exploited remain limited.
Who it affects: Financial services firms and large technology companies building AI products are HiddenLayer’s biggest customer sectors, Sestito said. The company also has contracts with the Department of Defense and the intelligence community. Sestito told TechCrunch that annual recurring revenue grew more than tenfold over the past year to tens of millions of dollars, with over 90% of that growth coming from new customers. He did not disclose an exact revenue figure. For business teams, the company’s expanded scope puts protection of operating AI systems and the models they adopt at the center of the purchasing decision.
What to watch: Competition and product overlap remain important considerations. TechCrunch reported that Noma and Zenity have each raised more than $100 million to address adjacent or overlapping security needs. Sestito acknowledged that some capabilities HiddenLayer sells could eventually be bundled into platforms from Microsoft, OpenAI and AWS, although he expects those platforms to emphasize governance features such as discovery, identity and policy controls. Buyers evaluating specialist AI security vendors should compare their coverage with protections already available from model providers and existing security platforms.
Our take
AI security deserves a dedicated place in deployment planning. Evaluate specialist vendors against the protections already available from model providers and existing security platforms.