Microsoft 365 E7 and the new shape of AI licensing

Bottom line. Microsoft brought Microsoft 365 E7 to general availability on 1 May 2026 at $99 per user per month, folding E5, Microsoft 365 Copilot and Agent 365 into a single tier, and on 1 Jul 2026 it raises the E3 and E5 base prices. The effect is structural, not promotional: the enterprise AI-licensing decision now lands on the renewal table as a tier choice, and a customer who declines Copilot still pays more for the base.

Microsoft’s 9 Mar 2026 announcement gave the enterprise software market a new top SKU. Microsoft 365 E7, branded the Frontier Suite, reached general availability on 1 May 2026 at $99 per user per month and unifies Microsoft 365 E5, Microsoft 365 Copilot and the new Agent 365 control plane into one licence, with Microsoft Entra Suite and the advanced Defender, Intune and Purview capabilities folded in. Agent 365, the layer for registering and governing AI agents, is also sold on its own at $15 per user per month.

The number that travels with E7 is Microsoft’s own framing of its value:

“At $99 per user, E7 is priced below purchasing these capabilities a la carte.”

— Judson Althoff, CEO, Microsoft Commercial Business, in the 9 Mar 2026 announcement.

That statement is true and beside the point for most buyers. “Below a la carte” only describes a saving for an organisation that wants every component in the bundle. For the larger group that wants two or three of them, E7 is a mechanism to pay for the others.

What changes on 1 Jul 2026

The bundle did not arrive on its own. Microsoft’s 2026 licensing update raises the base prices of the two tiers most enterprises actually run, effective 1 Jul 2026.

Per-user figures reflect Microsoft’s annual-commitment pricing, drawn from the licensing update and the 9 Mar 2026 post.

Two facts sit underneath the table. First, the increases land on the base productivity tiers, not the AI add-ons, so they reach customers who have made no AI decision at all. Second, E7 at $99 is positioned a short step above an E5-plus-Copilot-plus-Agent-365 stack whose E5 floor itself rises on 1 Jul. The bundle and the base increase are two halves of the same move.

AI moves from add-on to tier

For two years the enterprise AI purchase was an add-on: buy your productivity suite, then decide separately whether to attach Copilot. E7 dissolves that separation. The AI capability becomes a property of the tier you license rather than a line item you can isolate and defer. This is not unique to Microsoft; it is the observable direction across the platform layer in 2026, where AI is increasingly bundled into the base contract instead of sold beside it, a pattern we track in the enterprise AI cost and ROI read and the agentic-AI cost-governance piece.

The consequence for procurement is that the lever you used to have, deferring the AI spend by simply not attaching it, narrows. With E7 the spend is the tier; with the base increase the spend rises even if you attach nothing.

What this means at the renewal table

For a CIO, the practical effect is that the Copilot yes-or-no question is now entangled with the tier question, and both surface at renewal. Three buyer positions fall out of the pricing:

• All-in on the Microsoft agent platform. If you intend to register and govern agents through Agent 365 and run Copilot broadly, E7 is the cleaner single line item, and Microsoft’s a-la-carte framing works in your favour.
• Copilot for a subset. If only some roles get Copilot, the per-seat split is the whole analysis. E7 applied across the workforce overpays for every seat that does not use the AI; an E5 base with targeted Copilot and Agent 365 attach is usually the lower bill, at the cost of more SKUs to manage.
• No Copilot this cycle. You still absorb the base increase on 1 Jul. The one lever left is timing: renewing before the effective date holds current E3 and E5 pricing until your next renewal. That comparison sits inside the same logic as the Agentforce-versus-Copilot pricing read for buyers weighing platforms against each other.

The move before 1 Jul

One unhedged recommendation: model the three positions above against your actual renewal date before 1 Jul 2026, segmented by seat, not by headline tier. The bundle is a reasonable buy for the all-in case and an expensive default for everyone else, and the difference is entirely in how many of your seats genuinely use all four components.

If you are not adopting Copilot this cycle, the licensing update is explicit that customers renewing before the effective date keep current pricing until renewal. That is the one concrete saving on the table, and it expires on a date, not a quarter.

Holding-up note

The primary claim of this piece (that E7 plus the 1 Jul base increase restructures Microsoft 365 economics so the AI-licensing decision becomes a platform-tier decision taken at renewal, and that customers who decline Copilot still pay more) is on a 45-day review cadence, short because pricing ages fast and the base increase only takes effect on 1 Jul 2026. Three kinds of evidence would move the verdict: Microsoft reversing or deferring the base increases; Microsoft unbundling Copilot pricing below the E7 delta in a way that re-opens the add-on path; or adoption data showing E7 is the lower bill for the median customer, not only the all-in one. The Holding-up record for AM-202 captures what changes, dated. Pricing figures are from Microsoft’s own pages as of 5 Jun 2026.