Governance Just Got a Category

Gartner has published its inaugural Magic Quadrant for AI Governance Platforms. The category now has a name, a set of named vendors, and a formal place on every board’s transformation agenda. AI governance has moved from afterthought to required capability.

That formalisation is overdue. With EU AI Act obligations arriving in phases and the US Senate advancing the AI AGENT Act, the pressure to evidence how AI decisions are made and controlled is no longer theoretical. A Magic Quadrant tells you the market now agrees: enterprises must govern their AI.

But a quadrant of tools raises a prior question the quadrant itself does not answer. Govern based on what?

What Governance Tooling Assumes You Already Have

AI governance platforms monitor, control, and evidence. They watch model behaviour, enforce policy, and produce the audit record. Each of these functions operates over something: the processes AI runs inside. That is the assumption baked into the category. It assumes those processes are articulated, current, and visible.

Most enterprises cannot meet that assumption. They govern on top of undocumented process reality: workflows scattered across systems, tribal knowledge, and procedure documents that describe intent rather than execution. The governance platform sees the AI. It does not see the process the AI is embedded in.

This is not a tooling flaw. It is a sequencing flaw. A governance platform layered over invisible processes governs a black box with a dashboard attached. It reports on the parts it can observe and stays silent on the parts it cannot. For an AI Director, that silence is the risk. For a Risk Manager, it is the gap between a documented control and an executed one.

The principle is simple, and it holds across every governance investment. You cannot govern what you cannot see. Governance platforms provide the framework for control. They do not provide the visibility that framework depends on. That has to come from somewhere else, and it has to come first.

Read why process intelligence is the prerequisite for AI.

Tesco Got the Order Right: Governance, Process, Pragmatism

When Tesco set out its approach to AI readiness, it named three pillars: governance, process, and pragmatism. A blue-chip enterprise put process alongside governance as a foundational element, not a downstream detail. That ordering is instructive.

Governance depends on process visibility because a control cannot be evidenced against a workflow no one can see. Pragmatism depends on it too, because you cannot make sensible decisions about where AI belongs until you know how the work actually runs today. Both pillars rest on the same foundation. Process-first is not a vendor slogan. It is how mature enterprises operate when the regulatory stakes are real.

And the stakes are rising. The EU AI Act requires demonstrable governance over high-risk AI systems. The AI AGENT Act signals that agentic AI will face the same scrutiny. In both cases, the obligation is to show what happened, in which process, under whose authority. That is a demand for the executive view of how the organisation runs, and for risk leaders to evidence control effectiveness against real execution rather than a sample. Neither is possible without process visibility beneath the governance layer.

The Prerequisite Layer Beneath the Quadrant

IGX360 Insights is not an AI governance platform. It does not appear in the quadrant and does not compete with the vendors that do. It sits upstream of them.

IGX360 Insights makes your processes visible. It articulates how work actually executes, enriches that reality with governance and control context, and turns a scattered operational picture into something a governance platform can act on. The governance tooling then has something real to govern, rather than a black box to monitor. This is the layer the quadrant assumes you already have, and most enterprises do not.

Process articulation is the foundation. Governance is the framework built on top of it. Get the sequence right and the governance platform delivers on its promise. Get it wrong and you are evidencing the parts you can see while the risk lives in the parts you cannot. To see how the intelligence layer works in practice, meet Iggy.

Foundation Before Framework

A governance platform bought without process visibility governs a black box. It produces confident reports about an operational reality no one has actually mapped. The Magic Quadrant tells you which tools are strong. It does not tell you whether your organisation is ready to use them.

The prerequisite is process articulation. Build that first, and every governance investment you make afterwards earns its return. Skip it, and you are buying a framework with no foundation to stand on.

Are you governing your AI, or governing a black box with a dashboard attached?

Build the foundation governance tooling assumes you already have. Request a diagnostic.