The intelligence layer emerges as the control plane for enterprise AI

Overview

As enterprises move from AI experimentation into full-scale production, the challenge has shifted from accessing models to managing the organizational context agents need to act reliably. Speaking at FinOps X 2026, Microsoft's Cyril Belikoff argued that companies need a single intelligence layer trained once on who they are and how they work. Microsoft's answer is Microsoft IQ, paired with governance tools like Agent 365 and controls embedded in GitHub Copilot.

Key Takeaways

The enterprise AI challenge has shifted from accessing models to managing organizational context.
Microsoft IQ is a unified intelligence layer that connects agents to data, workflows, and institutional knowledge without retraining each agent.
Token economics, governance, and data readiness must align before enterprises gain sustained value from agents.
Belikoff warned that without a structured platform, developers could unintentionally expose sensitive data.
Microsoft embeds governance controls inside GitHub Copilot for model selection, content safety, and security.
Agent 365 lets organizations register, monitor, and govern AI agents like human employees.

The intelligence layer emerges as the control plane for enterprise AI

The shift to an intelligence layer

The central challenge has moved beyond model access.

›Enterprises now must manage the organizational context agents need to act reliably.
›The pressure to govern costs, secure data, and maintain accountability is redefining AI architecture.
›Token economics, governance, and data readiness must align before enterprises gain sustained value.

Belikoff, vice president of commercial cloud and AI at Microsoft, said companies want an intelligence layer with context about who they are and how they work, trained once on their structure, documents, meetings, structured data, and business processes, and kept consistent for many years.

Microsoft IQ

Microsoft's answer is a unified intelligence layer.

›Microsoft IQ connects AI agents to an organization's data, workflows, and institutional knowledge.
›It avoids requiring every agent to be retrained from scratch.
›Microsoft helps customers turn their data into their company IQ.

Belikoff said the key question is what platform companies need to consolidate data so it is joined, clean, and ready for AI, warning that if data is not clean, you get garbage in and out.

Governance and guardrails

Governance is emerging as the companion to innovation.

›Belikoff compared the risk to the unchecked sprawl of SharePoint sites in earlier platform cycles.
›Without a structured data and AI platform, developers could unintentionally expose sensitive information.
›Microsoft embeds governance controls directly inside GitHub Copilot.

Those controls cover model selection, content safety, and security decisions, appearing in the developer's natural workflow so they can pick the right model and reduce hallucination.

Agent 365

Microsoft built a way to manage agents like employees.

›Agent 365 lets organizations register, monitor, and govern AI agents.
›It applies identity, access controls, and accountability from the start.
›The goal is to manage agents the same way companies manage human employees.

The enterprise context

The discussion took place at a FinOps event.

›Belikoff spoke with theCUBE at FinOps X 2026.
›The themes were token economics, governance, and data readiness.
›The aim is helping enterprises move AI from experimentation to full-scale deployment.

The article notes theCUBE is a paid media partner for the FinOps X event, and that sponsors do not have editorial control over the coverage.

Frequently Asked Questions

What is an intelligence layer in this context?

It is a layer that holds organizational context, trained once on who a company is, how it works, its documents, meetings, structured data, and business processes, kept consistent for many years.

What is Microsoft IQ?

Microsoft IQ is a unified intelligence layer that connects AI agents to an organization's data, workflows, and institutional knowledge without requiring every agent to be retrained from scratch.

Why does Belikoff stress data readiness?

He said that if a company's data is not clean, AI produces garbage in and out, so consolidating joined, clean, AI-ready data is essential before deploying agents.

How does Microsoft handle governance?

Microsoft embeds governance controls for model selection, content safety, and security inside GitHub Copilot, and uses Agent 365 to register, monitor, and govern agents with identity, access controls, and accountability.

What is Agent 365?

Agent 365 is a Microsoft offering that lets organizations register, monitor, and govern AI agents the same way they manage human employees.

Microsoft positions the intelligence layer as the control plane enterprises need to scale AI safely.