Microsoft is weighing a new top-end enterprise subscription, internally discussed as an “E7” tier for Microsoft 365, that would bundle its Copilot assistant with an AI agent offering referenced as “Agent 365” in reporting and could cost up to $99 per user per month. The discussions signal a potential shift in how the company could price AI labor inside its productivity suite, treating autonomous software agents less like features and more like billable workers. If launched, the tier would arrive alongside already confirmed price increases taking effect in summer 2026, raising the financial stakes for enterprise IT buyers.
What the E7 Tier Would Include
Sources familiar with the discussions say Microsoft is considering packaging Copilot and the new Agent 365 hub into a single premium offering that sits above the current E5 plan. The company is reportedly exploring both traditional per-seat licensing and consumption-based billing, a hybrid approach that would let organizations pay a flat rate for access while also metering the work their AI agents perform. The rumored price ceiling of $99 per user per month would represent a significant jump from the current E5 rate, which itself is about to rise.
This is not the first time Microsoft has tried to assemble such a bundle. Business Insider reports the company previously paused a similar effort, though details about why it shelved the plan or what has changed internally remain scarce. The renewed internal discussions could reflect demand signals from enterprise customers or competitive pressure to monetize AI agent capabilities, though Microsoft has not publicly explained its rationale. If E7 arrives, it would likely be framed as the default way for large organizations to access the most advanced automation features, rather than as a niche add-on for experimental teams.
Agents Already Billed Like Employees
The E7 concept does not emerge from thin air. Microsoft already runs a metered billing system for AI agent work through Copilot Studio, where businesses build and deploy autonomous agents that handle tasks like customer service, data retrieval, and workflow automation. Organizations can purchase tenant-wide capacity packs priced at $200 per pack per month, with each pack providing 25,000 Copilot Credits. Pay-as-you-go options also exist for teams that prefer variable spending over committed capacity. Usage is metered through Copilot Credits, meaning the more work agents perform, the more a company can pay under capacity or pay-as-you-go models.
This structure already mirrors how companies budget for human headcount: leaders estimate the volume of work, allocate resources, and pay based on output. The E7 tier would formalize that logic at the subscription level, folding agent capacity directly into the seat license rather than requiring a separate purchase. For large enterprises running hundreds of agents across departments, the difference between buying capacity packs a la carte and bundling them into a per-user subscription could simplify procurement but also lock organizations into higher baseline costs regardless of actual agent usage in a given month. That shift would move AI from a discretionary experiment into a fixed cost that finance teams must budget for years in advance.
Summer Price Hikes Set the Stage
The E7 discussion lands against a backdrop of confirmed pricing changes across the Microsoft 365 lineup. According to Microsoft’s licensing bulletin, commercial pricing updates take effect on July 1, 2026, with packaging changes beginning to roll out in June. The Microsoft 365 E5 suite will move from $57 to $60 per user per month, a modest increase on its own but one that compounds across large deployments. Existing customers will remain on current pricing until their agreements renew, giving them a temporary buffer before new contracts push them into the updated rate structure.
A $3 bump on E5 may look small, but it resets the baseline from which an E7 tier would be measured. If the top-end bundle lands near $99, the gap between Microsoft 365 E5 at $60 and the AI-loaded version would be roughly $39 per seat per month. For a company with 10,000 employees, choosing E7 over E5 would add nearly $4.7 million in annual licensing costs. That math forces a direct comparison: does the productivity gained from bundled AI agents justify spending that could otherwise fund dozens of human salaries, additional infrastructure, or other strategic initiatives that compete for the same budget line.
The Two-Speed Enterprise Problem
Most coverage of the E7 rumor has focused on sticker shock, but the more consequential question is structural. By packaging agent capacity into a premium subscription tier, Microsoft could end up with a system where the largest organizations get the deepest AI integration while smaller firms either cobble together cheaper alternatives or delay agent automation. Consumption-based pricing, in theory, scales with usage and keeps costs proportional. But bundling it into a high-floor subscription means companies must commit to a steep minimum before they see any return, effectively turning advanced automation into a luxury feature.
This dynamic could accelerate a split in enterprise technology adoption. Large firms with the budget to absorb $99 per seat would deploy agents across finance, HR, legal, and operations, compounding efficiency gains quarter over quarter as agents take on repetitive tasks and orchestrate complex workflows. Midmarket companies, already stretching to cover E5 licensing, would face a harder choice: either stay on lower tiers and risk falling behind in process automation, or divert funds from other priorities to chase AI-driven productivity. The per-seat model rewards scale: the more seats an organization buys, the more agent capacity it can justify, and the faster it can amortize the cost against measurable improvements. Smaller organizations lack that arithmetic advantage and may find it harder to justify the most expensive, agent-heavy bundles.
What Comes Next for Enterprise Buyers
Microsoft has not officially announced the E7 tier or provided a timeline for its release. The company’s history of pausing and reviving this type of bundle suggests internal pricing debates are ongoing, and the final structure could differ from what sources have described. Enterprise buyers should watch the June 2026 packaging rollout closely, since the way Microsoft restructures its existing tiers will signal how much room the company is leaving above E5 for a premium AI offering. Any new bundle that consolidates Copilot, an “Agent 365” offering described in reporting, and metered capacity into a single SKU would be a strong hint that E7, or something like it, is under active consideration.
The broader bet Microsoft is making is that enterprises will eventually treat AI agent costs the same way they treat payroll: as a necessary, recurring expense tied to output rather than a discretionary software line item. If that bet is correct, the E7 tier would not just be a new subscription. It would represent a fundamental change in how organizations think about digital labor, shifting AI from an optional add-on to a core part of the workforce. For CIOs and CFOs, the challenge over the next few years will be to build governance, budgeting models, and performance metrics that can evaluate whether a $99-per-seat world of embedded agents truly delivers more value than the human-centric systems it aims to augment.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
