What’s the latest with Copilot in PowerPoint?

So you may be the unwilling person in charge of deploying Copilot to all of your PowerPoint content creators, but just keeping track of what is going on –  Brand Kits, optimisation, who gets access, and what even is an OAL anyway – seems like another full-time job on top of your full-time job. And then we haven’t even gotten to the sticky subject of what you actually do once you know what’s being updated. You’ll be relieved to know we’re here to help.

Think of this as your local PowerPoint Copilot newsagent. A place to drop by for the latest headlines, useful updates and interesting things we’ve spotted along the way. We’ll gather what we’ve learned from updates, sift through the noise and bring the useful bits together for you.

Roll up, roll up! Get your Copilot news here. No need to scour every corner of the internet. We’ll help you keep up without AI keeping you up.

How to use this article

Use the links to jump down to what you’re interested in, read the latest, or make a big cup of tea and just read the whole thing (who knows what goss you might discover).

#1: November 2026

• The return of custom fonts to Brand Kits
• Template assessment is coming
• Tokens may be required

Why are we writing this update log?

#1: November 2026

Over the past few months, we’ve seen visual refreshes, new functionality, disappearing features and the occasional “Hang on, wasn’t there a button here yesterday?” moment. Things move fast. So, here are a few of the Copilot updates that have caught our eye recently, along with why we think they’re worth knowing about.

The return of custom fonts to Brand Kits

If, like most, keeping your presentations on brand is important, fonts are a pretty big piece of the puzzle. Within your Brand Kit, you can define the fonts Copilot should use for headings, subheadings and body text, giving it a little more guidance when creating your slides. Previously, the ‘Add fonts’ option allowed you to choose from a range of system fonts or upload your own custom font.

Then, for a while, that custom font option disappeared. That meant you could either choose an available system font or leave the font undefined and rely more heavily on your PowerPoint template, embedded fonts and steering notes to help point Copilot in the right direction.

But good news: custom fonts are back! We’ve spotted that the option to upload your own font file has returned, giving you another way to tell Copilot exactly what “on-brand” should look like, rather than leaving your template to do all the heavy lifting. Will it stay? We are unsure, but, as we’re quickly learning, things in the world of AI can change fast. We’ll be keeping an eagle eye on this one and will let you know if anything changes again.

Template assessment is coming

Uploading a template to Brand Kit currently involves a certain amount of trust. You’ve built your template carefully, but how well will Copilot actually work with it? It looks like Microsoft is about to create an extra step that will give you eyes on your uploaded file.

Microsoft has announced an upcoming automatic template assessment that will check key areas of your PowerPoint template for compatibility when it’s uploaded to your Brand Kit. Once the assessment is complete, you’ll be able to review the results and see suggested tweaks that could help your template work more effectively with Copilot.

The assessment looks at some important areas, although it doesn’t replace the more detailed inspection we carry out as part of our Copilot-readiness service, where we dig much deeper into how the template is built and how it behaves as a whole.

And there’s another interesting development. Microsoft has since updated its guidance to suggest that Copilot itself will be able to help fix some of the issues identified in the assessment, directly within the template.

Naturally, this has caught our attention. We’ll be putting the feature through its paces as soon as it becomes available: seeing what it spots, what it misses and, most importantly, how successfully Copilot can make those suggested fixes without creating new problems along the way. We’re intrigued. We’ll be testing. Watch this space.

Tokens may be required

Another change we’ve got our eye on is usage-based billing. Microsoft has announced plans to introduce usage-based billing (UBB) for access to certain advanced AI models and capabilities across Copilot Chat and Copilot in Word, Excel and PowerPoint.

In practice, this could mean that access to some of the newest or more advanced AI capabilities won’t simply be switched on for everyone. Where usage-based billing is enabled by an organisation, users may be able to request tokens from their administrator when they need access to these capabilities.

Though we don’t anticipate this directly affecting the general user, it does introduce another consideration for organisations that want to stay at the front of the AI queue. Access to the latest and most advanced models could increasingly come with additional usage costs, making the question of “How much will AI cost us?” a little harder to answer with one neat number.

This is one we’ll be watching closely. Over the coming weeks, we’ll be digging into how the token system works, which Copilot experiences it applies to, how administrators can control access and, most importantly, what it could mean for you in the real world.

Why are we writing this update log?

If you’re anything like me, your phone is on silent. In fact, it’s probably been that way for years. No real-time ‘bings’ announcing every new message, just those little notification bubbles quietly creeping upwards throughout the day. Then you unlock your phone, glance at your apps, and there it is: 52 unread messages in the group chat.

Your heart sinks a little.

Do you know the feeling? You open the chat and immediately wonder where on earth to start. Do you even have the time, or the energy, to catch up properly?

You realise you have two options.

Option one: Get comfy on the sofa, strap in and start scrolling. You’re committed now. You don’t want to miss any important group intel or, more importantly, some excellent gossip. So you work your way through, reply with the occasional ❤️ or, if you’re more of a dad-style replier, a dependable 👍. Eventually, you emerge fully-informed, but precious minutes of your life have disappeared.

Option two: Put the phone back down. The messages can wait. Except, while you get on with your day, that little notification bubble keeps silently ticking upwards. And sooner or later, someone will ask you a question that makes you realise you’ve missed some fairly crucial context. At which point, you can either admit you haven’t read the chat or take a chance on a quick “Yep, sounds good!” and hope you haven’t just volunteered to babysit everyone’s children while they try the swanky new bar around the corner.

But what if there were a third option?

AI is starting to feel a lot like one enormous group chat. Look away for five minutes and there’s a new feature, a new tool or a new way of doing something you’ve only just learnt how to do. You could spend all day trying to keep up and still feel like you’ve missed something important.

The good news is, you don’t have to read all 52 messages yourself. That’s what we’re hoping to help with. At BrightCarbon, we’re constantly learning, testing and experimenting with AI. Instead of keeping those discoveries to ourselves, we want to share what we learn.

The AI group chat isn’t getting any quieter. Drop into our newsagents, grab the headlines and keep an eye on our posts, we’ll help you stay caught up, without having to read all piece together the parts of the story yourself.