Ultra Security Emerges From Stealth With $10 Million Seed for Agentic Security

Ultra Security emerged from stealth on October 21, 2026, with a $10 million seed round to introduce an agentic security platform that supplies real-time visibility, guardrails, governance, observability, and anomaly detection for autonomous AI agents. The company, based in Birmingham, Michigan, positions the product between agents and MCP servers to intercept tool calls, data reads, and prompt requests before execution.

Ultra Security Platform Capabilities and Certifications

The platform enforces policies written in plain language. Administrators can block credentials, personally identifiable information, or protected health information and prevent data movement across connectors. It supplies governance over agents, connectors, and tools while maintaining logs and traces that include AI-native anomaly detection. The system operates through a universal connectivity layer that supports multiple vendors and models without modification.

Processing defaults to local-first execution so audit trails remain on device when required, though a remote-hosted option exists for teams that need centralized management. Ultra Security has already obtained SOC 2 Type II, ISO 27001, and ISO 42001 certifications along with AARM Extended conformance. It aligns with the NIST AI Risk Management Framework and Agentic Trust Controls and has completed a FedRAMP 20x Class B audit while pursuing full authorization.

Early design partners include The Martec Group, which uses the tool to manage internal data access through agents connected to SharePoint, Outlook, and Slack. The platform automatically identifies agents and connectors such as Claude, Cursor, and GitHub and applies policies without manual configuration for each integration.

Founder Background and Funding Details

Chase Lee founded Ultra Security in 2025 after serving as CEO of Trustpage, which Vanta acquired, and later as VP of New Product at Vanta. He identified the need while observing that agents now receive direct access to tools and systems through MCP connectors, yet legacy controls such as web application firewalls and API gateways were built for static traffic rather than live agent sessions.

The seed round closed with participation from Ludlow Ventures, Hidden Capital, Entrée Capital, Detroit Venture Partners, firedrop, FDVC, and multiple angels from the security and SaaS sectors. No single lead investor was named in the announcement. Proceeds will support continued platform development and initial go-to-market efforts across organizations at different stages of AI adoption.

Access to the platform is currently available through a request form on ultra.security. The company states that the solution applies to both large enterprises and smaller teams regardless of current AI maturity level, focusing on the gap between rapid agent deployment and existing security tooling.