Microsoft’s latest changes are here. What do they mean for you?

Published October 8 2026

If you’ve opened LinkedIn lately, you’ve probably seen five different takes on what Microsoft’s latest announcements mean. Some are calling it a revolution. Others are calling it hype.

As usual, the truth probably sits somewhere in the middle.

Between the introduction of new Copilot experiences such as Home, Code and Autopilot, alongside updates to Microsoft 365 packaging and AI pricing models, there’s a lot to take in. Believe me, I know. Keeping up with Microsoft’s AI roadmap can sometimes feel like a full-time job.

Beneath the product names, licensing updates and launch-day excitement, however, sits a bigger story. Microsoft’s recent announcements suggest a future where AI becomes more deeply embedded into how work gets done, not just how individual tasks are completed.

The value of AI is evolving beyond prompts – this is one of the themes I took away from Microsoft’s announcements is that Copilot appears to be evolving from a tool that helps generate content and retrieve information into an experience that can execute tasks and support larger pieces of work.

For the past few years, many of us have experienced Copilot one interaction at a time:

• Draft this email
• Summarise that meeting
• Help me rewrite this proposal

Those experiences remain valuable today, but Microsoft’s latest announcements point towards something even more valuable.

The introduction of Home provides a central place to start work. Code creates a more specialised experience for developers. Autopilot introduces new possibilities around AI-assisted workflows and task orchestration.

What excites me isn’t any single feature. It’s what these experiences may represent collectively. My interpretation is that Microsoft is moving towards experiences that can retain more context, support longer-running work and reduce some of the coordination effort that sits between an idea and an outcome.

For example, imagine your Marketing Manager asking Copilot to prepare the first version of a campaign plan, gathering relevant information, structuring deliverables and highlighting decisions that still require your input.

Or imagine a frontline healthcare worker taking a large collection of client notes and having them organised into a consistent format before you begin working with them.

To be clear, these are examples of how I believe these capabilities could be applied in practice, rather than announced Microsoft scenarios. But they illustrate what many organisations are looking for: less time finding information and more time reviewing, refining and making decisions.

At Engage Squared, we’re already seeing elements of this today. Our marketing team recently used Copilot Cowork to migrate and shorten 36 customer stories into a new presentation-ready one-page format in around 15 minutes. That’s not a hypothetical productivity claim. That’s time handed directly back to the team.

Context is the game-changer

One of the developments I’m personally most excited about isn’t necessarily the flashiest. It’s context. AI becomes significantly more useful when it understands the work around you: your role, the people you collaborate with, the projects you’re involved in and the type of output you’re trying to create.

I’ve already noticed this in my own experience. When I use Copilot, the responses often reflect the fact that I work in marketing. The outputs it generates are usually different from what it might provide to a consultant, developer or executive. That additional context can make a surprisingly large difference to the quality of the starting point.

My view is that, over time, people’s expectations of workplace AI will continue to evolve. We’ll become less willing to explain everything from scratch and more likely to expect AI tools to understand enough context to help us get moving quickly, while still allowing us to remain in control.

Why Autopilot has everyone’s attention

The word “Autopilot” naturally attracts attention and probably a little nervousness too.

But I don’t see the opportunity as removing people from work. It’s about reducing the effort needed to manage the steps in between. Think about customer onboarding. Information has to be collected, checked and routed. Documents have to be prepared, notifications sent and approvals tracked.

Microsoft keeps investing in agents, orchestration and workflow. That suggests more of these structured activities can be coordinated with AI, while people stay responsible for judgement, creativity, approvals and decisions.

Not everything should be automated. But there’s plenty of busywork that shouldn’t need so much human effort.

What organisations should know about Microsoft’s pricing updates

Alongside new feature announcements, Microsoft has also introduced pricing changes for some of its AI-enabled experiences.

This has generated plenty of discussion, and understandably so.

The easiest way to misread them is to assume everything is moving to consumption billing. It isn’t.

Per-user licences, including Microsoft 365 Copilot, remain the foundation. What’s new is that some AI experiences now draw on Copilot Credits, particularly agents and more advanced workloads. Copilot Credits is a consumption model for work where the amount of AI processing varies a lot.

In practice, organisations may now see a mix of:

• Per-user licences for Microsoft 365 and Microsoft 365 Copilot
• Usage-based charges for some agents, workflows and advanced AI experiences
• Pre-purchased Copilot Credit pools that can be allocated across AI workloads
• Pay-as-you-go options that scale with demand

The logic is simple: not every AI workload costs the same to run. A quick question-and-answer exchange needs far less processing than an agent that retrieves information, reasons across multiple data sources, takes action and returns a recommendation. Think of it as the difference between a subscription and a utility bill. It isn’t unique to Microsoft either, since consumption pricing has been standard across cloud services for years.

What’s changing and what’s staying the same?

Key takeaways

• Traditional per-user licensing isn’t going away, but some AI experiences will be priced on usage.
• Not every user needs access to every AI capability.
• Visibility and governance matter more as AI usage grows.
• Consumption pricing can tie costs more directly to specific workloads and outcomes.
• Adoption, business alignment and change management are still what turn AI investment into value.

Successful AI programs won’t be measured by the number of licences assigned or agents deployed. They’ll be measured by the outcomes they deliver.

The bigger conversation isn’t cost. It’s value.

Ready to get started?

Keeping up with the technology is one thing but working out what is actually useful for your people, where to start, and how to scale it well is another. At Engage Squared, we help organisations move through that whole journey. That might mean starting small with a Copilot pilot, helping your teams build confidence around what tool, what model and when, or improving adoption so people are getting more than the basics from the licences you already have.

From there, we can help you identify the use cases worth pursuing, prepare your data and governance foundations, design and build Copilot agents and workflows, and work out where experiences like Cowork, Scout and Copilot Studio could genuinely add value.

The right starting point will look different for every organisation. The important thing is starting with the work your people are actually trying to get done, then choosing the technology that makes sense around it.

Ready to get started? Talk to our team about where Microsoft AI could make the biggest difference in your day-to-day.

Get in touch