June 10, 2026. On June 9, KPMG and Microsoft announced that KPMG will deploy Microsoft Agent 365 across its global network and put Microsoft 365 Copilot in the hands of all 276,000 of its professionals. Strip away the enterprise scale, and the message for every operator is the same: the hard part of AI agents is no longer building them, it is governing them.
What was announced
- Agent 365 will run KPMG's agents. KPMG will use Microsoft Agent 365 to deploy, manage, monitor, and update AI agents across its organization, with centralized visibility and control. The tooling plugs into KPMG Workbench, its multi agent platform built on Azure AI Foundry, and its Clara audit platform.
- Copilot goes to the whole workforce. Two years after a first rollout, KPMG is expanding Microsoft 365 Copilot to more than 276,000 professionals across 138 countries, on a platform that supports multiple models and protects the firm's data.
- Governance is the headline, not the model. The stated goal is to move clients from isolated pilots to organization wide deployment with security, identity, and lifecycle controls built in. Microsoft now classifies KPMG as a Frontier Firm, its label for companies rebuilding around AI.
What it means for operators
You do not need 276,000 employees for this to apply. The moment you let an AI agent send email, update a CRM, or touch customer data, you inherit the same questions a global firm faces: which agent did that, what is it allowed to do, who owns it, and can you switch it off. Most small teams deploy agents with none of those answers, then get surprised when one goes off the rails. A little governance, clear ownership, access limits, and a log of what each agent does, is what separates a reliable automation from a liability.
That discipline is exactly how we deploy agents for clients. Our AI automation builds include guardrails, access controls, and monitoring by default, and our AI automation agency can stand up governed, production grade agents for your team without the enterprise price tag. If you want hands on engineering help, you can also hire an AI engineer from us.
Frequently Asked Questions
Microsoft Agent 365 is a platform for deploying, managing, monitoring, and securing AI agents across an organization, with centralized visibility, identity, and lifecycle controls. It is designed to help companies move from isolated AI pilots to governed, organization wide agent deployments.
On June 9, 2026, KPMG said it will adopt Microsoft Agent 365 to govern AI agents across its global network and expand Microsoft 365 Copilot to all of its more than 276,000 professionals in 138 countries, integrating the tooling with its KPMG Workbench and Clara platforms.
Once an agent can act, by sending messages, updating records, or accessing data, you need to know which agent did what, what it is permitted to do, who owns it, and how to stop it. Governance provides that visibility and control, which is increasingly the gating factor for putting agents into production safely.
Start small but disciplined. Give every agent a clear owner, limit its access to only the systems it needs, log what it does, and keep a way to pause it. These lightweight controls deliver most of the safety benefit that large firms get from enterprise platforms, without the enterprise cost.