Chase Lee sold his last security company, Trustpage, to Vanta. His new one, Ultra Security, is a $10 million bet that the next big attack surface won’t be human. It will be autonomous.
Founded in 2025 and emerging from stealth in October 2026, the Detroit-based startup is building what it calls an AI-native security layer. The product sits between AI agents and the Model Context Protocol (MCP) servers, APIs, and connectors they use, aiming to be the control point for every tool call, data read, and prompt request [The Cincinnati Enquirer, October 2026]. In a market racing to connect agents to internal data, Lee is betting companies will pay to police that connection in real time.
The Runtime Enforcement Wedge
Ultra’s core proposition is intercepting actions, not just logging them. Conventional security tools monitor user activity. Ultra is built for a world where the ‘user’ is an AI agent with programmatic access to sensitive systems. Its architecture deploys a customer-deployed enforcement point that authenticates and inspects every agentic and MCP action as it happens [FedRAMP Marketplace, retrieved 2026].
The platform’s stated controls are designed for this new trust model. According to the company, customers can configure policies to block access to secrets, redact personally identifiable information, protect protected health information, rate-limit potential abuse, and prevent data from one connector from being written to another [The Cincinnati Enquirer, October 2026]. The goal is to provide a tamper-evident audit log and enforce guardrails defined once, automatically, across every agent action [Jamf Marketplace, retrieved 2026].
A Team Built on Compliance DNA
The founding team’s background is a central part of the pitch. Lee and colleagues hail from Vanta and Secureframe, companies that built businesses automating security and compliance audits for cloud infrastructure [Ultra Security, retrieved 2026]. Their collective experience is in selling ‘trust as a service’ to enterprises,a relevant playbook for convincing technical buyers that agentic workflows need a new kind of oversight.
Lee’s exit with Trustpage provides a track record investors could underwrite. The seed round, announced at emergence, drew capital from a mix of venture firms and angels known for early-stage software bets.
| Investor | Known For / Notable Focus |
|---|---|
| Ludlow Ventures | Early-stage consumer and SaaS investing. |
| Hidden Capital | Early-stage B2B software. |
| Entrée Capital | Global early-stage venture firm. |
| Detroit Venture Partners | Detroit-focused technology investing. |
| firedrop | Early-stage venture firm. |
| FDVC | Fintech and software venture capital. |
| Table: Lead investors in Ultra Security’s $10 million October 2026 seed round [Business Wire, October 2026]. |
The MCP Tailwind
Ultra’s bet is tightly coupled to the adoption of the Model Context Protocol. MCP is an open protocol that standardizes how AI models connect to data sources and tools, championed by companies like Anthropic. As developers build more agents that use MCP connectors to interact with internal systems, the need for a standardized security layer grows. Ultra positions itself as an early contributor to frameworks like AARM and the Agentic Trust Control Framework, aligning its product roadmap with the ecosystem’s evolution [Ultra Security, retrieved 2026].
The initial traction signals are early but pointed. The company is hiring for two remote, senior engineering roles focused on applied AI and security infrastructure, indicating a build phase focused on core technology [Ultra Security, October 2026]. A quote on its homepage from Maneesh Apte, Co-Founder and CEO of Parker, suggests engagement with early design partners, though the commercial scope is not disclosed [Ultra Security, retrieved 2026].
The Inline Interception Risk
For all its technical promise, Ultra’s model introduces a classic infrastructure sales challenge. Its security relies on being deployed inline, as a mandatory gateway for all agent traffic. This creates a single point of failure and potential performance bottleneck that engineering teams may resist. Selling runtime enforcement is harder than selling post-hoc monitoring; it requires deep integration into the development pipeline and a conviction that security must be proactive.
The competitive landscape is also unformed but inevitable. Established cloud security players and emerging AI infrastructure startups will likely develop adjacent capabilities. Ultra’s answer rests on being AI-native and MCP-first,building deeper, more granular controls for agentic workflows than a general-purpose tool could provide. Its early-mover focus on this specific protocol could be its wedge.
What to Watch in the Next Twelve Months
The $10 million seed provides an 18-24 month runway to prove the model. The next milestones are straightforward: land the first public reference customers, demonstrate performance at scale, and expand the policy library. The hiring of a GTM Engineer role suggests go-to-market experimentation will begin soon [Ultra Security, October 2026].
Success will be measured in deployments, not just logos. Can Ultra become the default security layer for companies rolling out Claude and Cursor agents at scale? The seed backers,Ludlow Ventures, Hidden Capital, Entrée Capital, Detroit Venture Partners, firedrop, and FDVC,are betting the team that navigated compliance can now navigate runtime [Business Wire, October 2026]. The question for enterprise security chiefs is whether the agent threat model is urgent enough to buy a dedicated solution today, or if it’s a problem for next year’s budget.
Sources
- [Business Wire, October 2026] Ultra Security Emerges from Stealth with $10M Seed Round to Secure Autonomous Runtimes | https://www.morningstar.com/news/business-wire/20261021802328/ultra-security-emerges-from-stealth-with-10m-seed-round-to-secure-autonomous-runtimes
- [The Cincinnati Enquirer, October 2026] Ultra Security Emerges from Stealth with $10M Seed Round to Secure Autonomous Runtimes | https://lifestyle.middletownlifemagazine.com/story/774466/ultra-security-emerges-from-stealth-with-10m-seed-round-to-secure-autonomous-runtimes/
- [Ultra Security, retrieved 2026] Ultra Security - Runtime security for AI agents | https://ultra.security/
- [Jamf Marketplace, retrieved 2026] Ultra Security listing | https://marketplace.jamf.com/details/ultra-security
- [FedRAMP Marketplace, retrieved 2026] Ultra Security listing | https://marketplace.fedramp.gov/listing/ultra-security
- [Ultra Security, October 2026] Careers - Ultra Security | https://ultra.security/careers