Workplace AI

Microsoft 365 E7 Is Now Generally Available: What the Frontier Suite Means for Your Organisation

Microsoft's first new enterprise licence tier since 2015 bundles E5, Copilot, Entra Suite, and Agent 365 into one platform at $99/user/month.

Workplace AI category

Microsoft 365 E7, branded as the Frontier Suite, became generally available on 1 May 2026. It is the first new enterprise licence tier Microsoft has introduced since E5 launched in 2015, and it represents a fairly significant shift in how Microsoft is packaging and positioning its AI and security products for large organisations.

If you are currently running E5 with Copilot, or evaluating how to scale AI across your organisation without juggling multiple separate licences, this one is worth understanding properly.

What is Microsoft 365 E7, exactly?

E7 is a single licence that bundles four products that many enterprise customers were already buying separately:

  • Microsoft 365 E5 (the existing security and productivity suite)
  • Microsoft 365 Copilot (the AI assistant integrated across Word, Excel, Outlook, Teams, and more)
  • Microsoft Entra Suite (the full identity and network access management stack, beyond the Entra ID included in E5)
  • Agent 365 (the new control plane for deploying, monitoring, and governing AI agents at scale)

The combined price is $99 per user per month. Buying those four components individually would cost around $117 per user per month, so the bundle saves roughly 15%. That saving only makes sense, of course, if you genuinely need all four components. More on that below.

Agent 365 is also available as a standalone licence at $15 per user per month for organisations that want the agent governance layer without upgrading their entire licence tier.

What is Agent 365 and why does it matter?

This is the new piece most worth paying attention to. As organisations move from one or two Copilot pilots to running dozens or hundreds of autonomous AI agents, the question of governance becomes critical fast. Agent 365 is designed to be the central control plane for that.

It gives IT and security teams the ability to observe, govern, and secure agents running across the organisation. That includes agents built in Copilot Studio, agents from third-party partners like ServiceNow, SAP, and Workday, and eventually agents running on other cloud platforms. Microsoft has announced a public preview of registry sync with AWS Bedrock and Google Cloud connections, which would let IT teams discover and apply basic lifecycle governance to agents regardless of where they are running.

From a security perspective, Microsoft Entra Conditional Access now extends Zero Trust policies to agents acting autonomously or on behalf of users. The same risk-based access evaluation that applies to human identities applies to agent identities. Starting in June 2026, Microsoft Defender will also provide asset context mapping for each agent, including the devices they run on, associated identities, and the cloud resources those identities can reach.

There is also a new Shadow AI page in the Microsoft 365 admin centre (currently in Frontier preview) to help identify and block unauthorised agents before they create compliance or security exposure.

What else is new in this wave?

E7’s GA arrived alongside a broader set of updates to the Copilot platform:

Work IQ is the intelligence layer that gives Copilot and agents a deeper understanding of your role, your work patterns, and your organisation’s context. It improves Copilot’s usefulness across Word, Excel, PowerPoint, and Outlook rather than treating every interaction as if it is starting from scratch.

Anthropic’s Claude Opus 4.8 is now available inside Microsoft 365 Copilot, giving users access to a second model for complex, multi-step work. This expands the model options beyond OpenAI and is particularly aimed at long-running tasks, document drafting, data analysis, and building presentations.

Copilot Cowork is now on iOS and Android. You can delegate work from your phone, pick it back up on your desktop, and keep tasks moving without losing context between devices.

What does this mean for you?

If you are an IT decision-maker at a large enterprise on E5 with Copilot: E7 is the natural next step if you are also planning to deploy agents at scale and need the full Entra Suite. Run the numbers for your seat count. At 15% cheaper than buying the components separately, the savings are real, but only if Agent 365 and the full Entra Suite are on your roadmap anyway.

If you are on E3 or a smaller tier: E7 is not a direct upgrade path from those tiers. Microsoft 365 Business Basic, Standard, and Premium are unaffected and remain available. E7 is aimed squarely at large enterprise deployments, typically 500 seats or more, where AI governance and identity management complexity justifies the investment.

If you are responsible for security or compliance: Agent 365 standalone at $15 per user per month is worth serious consideration even without the full E7 upgrade. The agent governance and Conditional Access for agents story is genuinely new capability, not repackaging of existing features.

If you are a Microsoft partner: From April 2026, E7 and Agent 365 are included as eligible workloads in CSP incentives, and Agent 365 is now part of Copilot and Power Accelerate engagements including Immersion Briefings and Proof of Concept work.

One licensing detail worth flagging

The May 2026 Product Terms update revised the definition of multiplexing to explicitly reference automation. If your organisation is deploying AI agents, bots, or automated workflows that interact with Microsoft 365 services, this change may introduce additional licensing considerations that were not previously relevant. It is worth a conversation with your licensing advisor before you scale agent deployments. RixMind has a useful breakdown of what the multiplexing update means in practice.

The broader picture

Microsoft’s 2026 Work Trend Index surveyed 20,000 workers across 10 countries and analysed trillions of anonymised Microsoft 365 signals. One finding stands out: organisational factors, including culture, manager support, and talent practices, account for 67% of reported AI impact. Individual mindset accounts for 32%. The technology is largely ready. The organisations often are not.

Only 19% of AI users in the research sit in what Microsoft calls the “Frontier” zone, where both individual capability and organisational readiness are high. Another 10% have strong individual AI capability but work in organisations that are not set up to use it.

That context matters for how you think about E7. The licence gives you the infrastructure to run enterprise-grade AI at scale. What it cannot do is change the culture or management practices that determine whether that infrastructure actually delivers value. The Frontier Suite is a meaningful step forward in how Microsoft packages and governs enterprise AI. Getting value from it is still a people and process challenge as much as a technology one.