Opsin Raises $21M for AI-Agent Security

Opsin Security has raised $21 million to monitor autonomous AI agents after deployment and detect when their behaviour drifts away from assigned purposes or permissions.

What happened

The financing was led by SYN Ventures and Race Capital. Opsin initially focused on preventing sensitive-data exposure through enterprise AI tools and is expanding toward broader agent governance. Its platform is designed to observe agent behaviour over time rather than inspect isolated prompts, helping security teams identify when an agent starts accessing unexpected systems, handling sensitive information or taking actions outside its intended role. The product is aimed at enterprises deploying agents across finance, CRM, coding and other internal systems.

Why it matters

Traditional security tools are designed around users, endpoints and network traffic. Autonomous agents create a new risk because software can independently perform sequences of actions across multiple systems. Understanding the agent's ongoing intent and behaviour can therefore matter more than evaluating one request at a time.

The bigger picture

Agent governance is emerging as a distinct cybersecurity category. Opsin's long-term differentiation will depend on detection accuracy, low false positives and integration with identity, access-control and monitoring systems already used by enterprises.