Enterprises deploying AI agents at scale now face a second problem: a crowded market of vendors selling ways to secure them. At least two dozen companies now offer some form of AI agent security, according to public Crunchbase and PitchBook profiles. Their products range from vetting the tools agents use to controlling what data agents can reach, detecting threats on devices running agents, and finding unapproved AI usage.
The promises sound similar across vendors - knowledge graphs, continuous monitoring, runtime security, tool access, MCP vetting. But the underlying approaches vary. Some focus on tool vetting, others on data access controls, and companies like CrowdStrike are building detection and response controls on the devices where agents run.
Reco, an AI security startup, announced Tuesday it raised $55 million in a Series B extension, bringing its total capital to $140 million. AT&T invested through its venture arm, along with Forestay and Quadrille Capital. The company said its valuation has more than doubled since its $30 million Series B in February, and co-founder and CEO Ofer Klein estimated it in the "high hundreds of millions."
The scale of unmanaged agents
Klein said the biggest change over the past year is that companies are building and deploying AI agents faster than they can track them. At one Fortune 100 customer, Reco's platform found 21,000 agents the company didn't know about. At a large financial services customer, the startup identified an agent set up by an ex-employee that could access Salesforce and share that data with a domain the company couldn't see.
"The market demand right now for agent security is not only about the agent itself; it's about the entire ecosystem end-to-end," Klein said.
The numbers from other vendors tell a similar story. HiddenLayer co-founder and CEO Chris Sestito said over 50 of his customers have AI agents in production touching critical systems and sensitive assets. Cymphony, another AI startup, said it found about 85,000 files that had become accessible to AI tools and agents at one U.S. public company.
Repositioning around agent ecosystems
Until last year, Reco mostly sold software to map and secure SaaS and AI platforms. The startup has now repositioned around a broader solution that uses a context graph to connect agents to apps, people, accounts, and permissions. Security teams can see what an agent can reach and cut off access it doesn't need.
Reco integrates with more than 280 apps, and Klein said new integrations can be added within days. The platform uses browser and network signals to find agents outside apps it connects to directly, and it has controls to inspect prompts and tool calls. Financial services companies account for about 40% of Reco's business, and the startup has more than 100 customers. Annual recurring revenue is in the double-digit millions of dollars, Klein said, and he expects it to triple this year.
For IT and security leaders evaluating agent security tools, understanding the scope of unmanaged agents inside their own organizations is a practical first step. Courses on AI Security Analytics can help security analysts build the skills to assess these risks. The funding will go toward hiring, sales, partnerships, and customer support.
Why this matters for IT and security professionals
If Reco's findings at a single Fortune 100 customer - 21,000 unknown agents - reflect what's happening at other large enterprises, security teams are already behind. The immediate task is not choosing a vendor but discovering what agents exist, what they can access, and who created them. An ex-employee's agent with live Salesforce access is a concrete example of the risk that comes from deploying agents faster than governance can keep up.
Your membership also unlocks:
