SailPoint completed its acquisition of Entro Security in June 2026, adding non-human identity and AI agent discovery to its platform. For most organizations the first useful step is not policy, it is discovery: finding the agents, MCP servers, service accounts and secrets already in use, then deciding who owns each one.
The biggest takeaway from SailPoint Navigate 2026 was the Entro Security acquisition. To get right to the point: based on what I learned at Navigate, SailPoint is offering proofs of concept at no cost, and getting one started should be an immediate priority for everyone. The terms are SailPoint's to set, so confirm the current offer before you plan around it.
What SailPoint Bought, and Why It Matters Now
SailPoint announced its intent to acquire Entro Security, a Tel Aviv-based company, on June 15, 2026, and closed the acquisition on June 29, 2026. SailPoint describes Entro as "a pioneer in non-human identity (NHI) and credentials security," and says Entro's NHI and credentials security solutions are available to SailPoint customers as standalone offerings. The product is now documented as SailPoint Entro. Financial terms were not disclosed.
Non-human identities are the service accounts, API keys, tokens and certificates that let software talk to software. AI agents add a new layer: every agent runs on credentials like these, and many reach their tools through Model Context Protocol (MCP) servers. On its agentic AI security product page, SailPoint says Entro provides "a complete agent and MCP inventory across your Cloud, SaaS, SDLC, and endpoints" and maps "every AI agent, NHI, and secret to a human owner." Machine credentials with standing access are exactly the kind of account privileged access programs exist to control.
Why Discovery Comes First
You cannot govern an identity you cannot see. Access reviews, least privilege, rotation and ownership all assume an inventory, and for non-human identities that inventory rarely exists in one place. SailPoint's closing announcement says Entro discovers over 1,200 types of secrets, tokens and certificates across CI/CD pipelines, codebases and container registries. The standard of care already expects the inventory: CIS Controls v8 Safeguard 5.5 calls for an inventory of service accounts that records, at a minimum, a department owner, a review date and a purpose, and for reviews at least quarterly to confirm every active service account is authorized.
As I understand it, the only catch is a legal agreement. The tool will detect, and then host, information about the non-human identities and agents in your environment, so it makes sense that SailPoint needs to protect itself. Have your security and privacy teams review it like any SaaS data term.
In a few conversations at the event, SailPoint representatives suggested the POC should really be paid, because in practice it does the detection legwork and helps organizations prioritize remediation. I agree it could be a paid engagement. But knowing most organizations are behind the eight ball, as the entire industry is, I'd bet a POC will be eye-opening. In some stories shared at the event, a POC was stopped partway through because a security leader grew concerned about their job after seeing the exposure, and security leaders were texting their CIO mid-POC to start the PO process. Those are secondhand, but they tell you what a first look at the non-human footprint can do.
Getting a handle on AI agents and MCP servers is strategic. Boards will find budget for Entro, and I don't see it coming out of existing security budgets. Yes, it's a security tool, but AI more broadly is a business enablement tool. It builds competitive advantage and helps organizations get more out of their resources and people. One CISO at the event explained that for the first time in his career, he had the board's full attention. He was the most important person in the room, because he was the one person who knew what needed to happen to enable this incredible technology safely.
What the Proof of Concept Involves
As I understand it, the POC is simple: you set up the tool, scan your environment, and complete the critical first step, discovery. From what I heard at Navigate, you don't need to own a SailPoint product to take part, and SailPoint IdentityIQ customers can absolutely take part. SailPoint's public announcement describes Entro as available to SailPoint customers as standalone offerings, so confirm eligibility for your situation with SailPoint. It doesn't matter where you are in your identity journey.
What Happens After Discovery
Discovery produces a list. The value is in the decisions about each entry:
- Owner. One accountable human for every agent, MCP server, service account and secret.
- Purpose. What it is for, in a sentence a reviewer can check.
- Access. The least access that purpose needs, with standing privilege removed where it is not.
- Review. A cadence, so the inventory stays true after the POC ends.
That is where discovery hands off to governance. Entro maps identities to owners; proving that ownership holds up, and building it into your review and remediation processes, is the work that follows. Keep that work fast. If getting an owner and access approved for a new agent takes weeks, teams will stand one up on credentials nobody reviews, and the inventory is stale again. Our AI Identity Governance Assessment is built for that step, and I cover what an AI governance program looks like after detection in a separate post. If you are earlier in the journey, an IAM assessment puts the findings in the context of your whole program.
The One Step to Take This Week
So I implore anyone reading this to get on the POC. This should be a today task. Get the ball rolling.