Reco Secures $55M to Tackle AI Agent Security Sprawl

The market for AI agent security is rapidly crowding, with dozens of vendors promising discovery, governance, and runtime protection for autonomous software agents.

Reco, formerly focused on SaaS and AI platform mapping, has pivoted to a broader context‑graph solution that links agents to applications, identities, and permissions. The startup announced a $55 million funding extension, building on a $30 million Series B from February. Investors include AT&T’s venture arm, Forestay, and Quadrille Capital, and the company’s valuation has more than doubled since the earlier round.

Early deployments illustrate the scale of the problem. At a Fortune 100 customer Reco uncovered 21,000 previously unknown agents, while a financial services client discovered an ex‑employee’s agent still accessing Salesforce data. Annual recurring revenue is in the double‑digit millions and expected to triple this year across more than 100 customers, 40% of which are in financial services.

Why it matters for GPU / AI infrastructure

• Agent proliferation drives demand for high‑performance GPU clusters to run inference and training workloads.
• Security gaps can expose sensitive model weights and data stored on GPU‑accelerated platforms.
• Unified visibility across agents, apps, and hardware helps operators allocate compute resources safely and efficiently.

As enterprises accelerate agent deployment, integrated security that spans software and the underlying GPU infrastructure will become a prerequisite for scalable AI operations.