HiddenLayer, an Austin-based AI security company, closed a $100 million Series B funding round led by Delta-v Capital, with participation from Microsoft's M12, Booz Allen Ventures, Ten Eleven Ventures, and Morgan Stanley, according to a September 2, 2026 report by TechCrunch. The financing brings together capital from cloud hyperscalers, defense contractors, and traditional finance behind a single platform focused on protecting AI systems from attacks. Co-founder and CEO Chris Sestito told TechCrunch that the company's annual recurring revenue expanded more than tenfold over the past twelve months into what he characterized as the "tens of millions," a milestone powered overwhelmingly by fresh customer acquisition rather than expansion of existing accounts.
The report states that over 90% of HiddenLayer's growth came from more than 50 net-new customers spanning banking, insurance, pharmaceuticals, airlines, and U.S. defense and intelligence agencies. One unnamed client is described as a "leading frontier model provider" serving more than 700 million weekly users. Named customer categories include financial services firms, large technology companies building AI products, the Department of Defense, and the intelligence community, according to the coverage. The company also secured contracts with the Department of Energy's $60 million Prometheus initiative and the Missile Defense Agency's SHIELD program, which the report notes add federally validated revenue streams to HiddenLayer's commercial base and confirm the platform's credibility across two separate government buyer segments. TechCrunch did not disclose the round's valuation or identify specific enterprise customers by name.
Sestito framed AI security as a consistent challenge regardless of deployment model, telling TechCrunch that "inference is still inference" whether applied to traditional machine learning, generative AI, or agentic workflows, meaning much of the company's technology remains applicable across formats. The platform addresses discovery, runtime protection, attack simulation, and supply-chain security, targeting threats such as prompt injection, agent manipulation, and malicious tool use. On the supply-chain front, Sestito explained that HiddenLayer parses and scans roughly 50 different AI file frameworks to verify that open-source and open-weight models are exactly what they claim to be, describing the risk as "hidden models inside of models." Additional coverage highlighted that HiddenLayer's method doesn't require access to raw training data or model weights, and that the company holds a 39-patent intellectual property portfolio while contributing to standards bodies including CISA, MITRE, NIST, and OWASP.
HiddenLayer plans to allocate the new capital toward sales hiring, engineering teams, and expansion into Europe, the Middle East, and Africa, the report states. The financing arrives amid a cluster of AI security developments, including CrowdStrike's SafeMind launch and OpenAI's announcement that its Astra system crossed a critical cybersecurity threshold. One investor quote surfaced in alternative coverage but not in the TechCrunch piece suggests that AI requires its own distinct category of protection, reinforcing the view that runtime security for AI systems is becoming mandatory enterprise infrastructure rather than an optional add-on. The investor consortium—spanning a defense-focused venture firm, a major cloud provider's investment arm, a traditional investment bank, and a government contractor—signals that organizations across sectors now view AI security as foundational rather than experimental. If enterprise adoption of agentic AI accelerates as forecasted, the race to protect those systems at runtime may define competitive advantage as sharply as model performance itself.