Opsin, an AI governance and security platform designed to detect and contain AI agent misbehavior, announced on October 15, 2026, that it has raised US$21 million in funding. The round was led by SYN Ventures and Race Capital. The investment will help enterprises securely adopt AI agents at scale as incidents involving rogue agents become more common.
The Risks of Goal-Oriented AI Agents
Goal-oriented AI agents are systems designed to independently take a series of actions to complete an assigned task instead of simply responding to individual prompts. Because these agents pursue goals, they can take extreme or unauthorized actions that create security risks, expose sensitive data, bypass controls, or disrupt critical systems. Opsin's analysis found that 93% of misaligned tools connected to agents had full, unrestricted access, while only 1 in 200 required human approval before acting.
Platform Capabilities and Security Context
Opsin gives security teams visibility, context, and control over AI agents built by employees across Microsoft Copilot, Claude, ChatGPT Enterprise, Gemini, and other enterprise-approved tools. The platform connects an agent's purpose, permissions, data access, tools, and activity to build a complete picture of how and why it operates. Security teams can identify when access or behavior falls outside what is appropriate, prioritize significant risks, and address issues before they become security incidents.
Executive Perspective on Agent Security
James Pham, co-founder and CEO of Opsin, explained that enterprises are moving toward a world where employees supervise teams of agents that work across data, knowledge, and systems with increasing authority to act. He noted that knowing what an agent has permission to access is no longer enough. Security teams need to understand what the agent is trying to do, who is behind it, what data it is using, and how its actions connect to the broader workflow.
Founding Team and Investor Support
Opsin was founded by MIT alumnus James Pham, cybersecurity veteran Oz Wasserman, and longtime engineering leader Jeremy Mailen. Pham and Wasserman first worked together at Abnormal Security, where they saw the challenges of securing increasingly intelligent systems. Alfred Chuang, General Partner at Race Capital, said the team has the security experience and technical depth to solve a problem that becomes more urgent with every agent deployed.
Enterprise Traction and Expansion Plans
The company already serves nearly 20 enterprise customers, including Culligan, Barry-Wehmiller, Wellstar Health System, and UiPath across healthcare, financial services, manufacturing, and other regulated sectors. Glenn Chisholm, Partner at SYN Ventures, said Opsin provides enterprises the confidence to embrace AI agents without opening the door to uncontained risks. With the new funding, Opsin will expand its platform to provide the visibility and control enterprises need to ensure their AI agents behave as intended.
A Security Foundation for the Future
Opsin is an AI-native governance and security platform built to help enterprises securely adopt AI agents at scale across workforce approved AI platforms. It connects identity, agent, data, tool, behavioral, and intent context to give security teams visibility into what AI agents can access, what they are intended to do, and when behavior deviates. The platform helps organizations discover AI agents, identify oversharing and excessive access, prioritize risk, and enforce security controls across enterprise AI environments.
As AI agents take on greater responsibility and become more embedded in how work gets done, the need for dedicated governance and security controls will continue to grow. Opsin is building for a future in which teams of agents act on an enterprise's collective knowledge, data, and processes. Its platform aims to provide the security foundation beneath that shift, protecting businesses as agents work across applications, processes, and platforms.