Sentra raises $150M led by Capital One Ventures for AI agent data security

Sentra, a data security startup that tracks what AI agents are allowed to touch inside corporate networks, has landed a $150 million infusion led by Capital One Ventures, with a Capital One representative set to join its board. Hoodline reported the round on November 4.

The company is co-headquartered in New York City and Tel Aviv. Sentra says the money will fund broader visibility into AI agents, copilots and models, along with tighter controls over the sensitive data those systems can reach. Existing investors Bessemer Venture Partners, Zeev Ventures, Standard Investments, Munich Re Ventures and Key1 Capital participated alongside Capital One Ventures.

Accounts of the round differ slightly in framing: a PR Newswire release describes the $150 million as a strategic capital investment, while SiliconANGLE reported it as a Series C. SiliconANGLE also reported that Sentra's total funding since its 2021 launch now exceeds $250 million, though the company has not disclosed a valuation.

Sentra says the past year brought more than 300% year-over-year growth, a surge it attributes to enterprises racing to deploy AI agents without fully understanding what those agents can access. The platform finds and classifies data across cloud, SaaS, data warehouse, knowledge store and on-premises file share environments, then maps which assets are reachable by human users, AI agents and service accounts alike.

According to SecurityWeek's reporting, the system enforces least-privilege policies, flags unusual access to sensitive data and automates remediation by revoking access and blocking risk once detected. It also generates reports on agent-accessible data framed with business context, and the company says it analyses information where it already lives rather than moving it elsewhere, so sensitive data never leaves the customer's environment.

Co-founder and CEO Yoav Regev frames the investment as a response to a structural problem: traditional perimeter security was not built for agents operating at machine speed. His argument is that securing the data agents can reach lets organisations deploy AI faster while staying in control, a balance the company is betting enterprises will pay to maintain.

The market is getting crowded quickly. A scan of public Crunchbase and PitchBook profiles by TechCrunch turned up at least two dozen companies now selling some form of AI agent security. Reco raised $55 million, Zenity closed a $125 million Series A-plus round led by Norwest in August, and Cyera's valuation reportedly quadrupled to $12 billion after a $600 million raise in June. Consolidation is picking up too, with CrowdStrike planning to acquire Fiddler AI and Dreadnode.

Survey data explains the investor appetite. A Cloud Security Alliance survey found that 82% of organisations had discovered previously unknown AI agents in their environments within the past year, and 65% reported at least one agent-related security incident; 61% of those incidents involved data exposure, 43% caused operational disruption and 35% led to financial losses. Only 21% of respondents said they had formal processes to decommission agents once they were no longer needed.

Sentra says it will use the funding to expand its go-to-market organisation, grow brand and marketing teams and scale field and channel programmes, and it plans to enter new geographic regions, though neither the specific markets nor a timeline has been disclosed.