Insights: the European Collaboration Summit 2026 was driven by AI and "SharePoint 3.0"
- Aron Beyens
- 15 hours ago
- 6 min read
European Collaboration Summit 2026 made one thing clear: we are entering a third era of SharePoint and Microsoft 365. This is an AI‑driven platform where content, Copilot, and agents operate as a single ecosystem rather than separate tools.
The scale of this shift is not always easy to grasp, but #ECS2026 reflected it clearly, with more sessions, speakers, and attendees than ever before. Where classic SharePoint focused on document storage, and modern SharePoint improved intranet usability, this new phase goes further. SharePoint and Teams are becoming the knowledge backbone for AI and agents.
For us at Flow, this shift feels both familiar and validating. For years, we have emphasised the importance of good content and solid information architecture. ECS confirmed that these foundations are no longer optional. They determine whether AI is trustworthy and useful. If content is messy, outdated, or hard to find, AI will not fix that. It will simply scale the chaos.
Let's illustrate this with a few lessons learned.
Lesson 1 – SharePoint is now an AI knowledge platform
Across keynotes and sessions, one message kept returning: SharePoint is no longer just a document library or intranet. It is the primary source from which Microsoft 365 Copilot and new agent experiences learn how your organisation works and what it knows. This means, you're going to need an "AI‑first" mindset:
AI-ready content matters SharePoint now includes capabilities to create, enrich, and organise content so Copilot and agents can give better answers (for example by flagging outdated pages, broken links, and low-quality content). It's called - this couldn't be more creative - "AI in SharePoint".
Permissions still rule Copilot and SharePoint‑based agents respect the existing permissions. Even when users ask natural‑language questions, they only see content they are authorised to access.
Structure becomes fuel
Clear sites, libraries, metadata, and well‑designed pages make it easier for AI to understand and reuse your content. This is not classic or modern SharePoint. It is effectively a new version. "SharePoint 3.0".
Lesson 2 – Copilot is your best friend, or at least it's trying really hard to be: Agents automate the “routine asks”
Copilot agents can take over many recurring, simple information requests. Using Microsoft 365 content, such as emails, SharePoint documents, Teams files, they can answer questions, summarise information, and guide users through standard processes. With "AI in SharePoint" you'll now have new options to implement and use Copilot in SharePoint and Teams. What this means in practice:
From FAQ pages to conversational help Instead of searching through PDFs or FAQ pages, employees can simply ask, “How do I request a new laptop?” and receive an answer based on existing documentation. SharePoint even allows FAQ pages to be generated using a dedicated FAQ page component.
Reusable knowledge Well-designed agents can be shared across Teams, SharePoint, and Copilot Chat providing consistent answers wherever users work. With SharePoint Skills, agents can also perform actions directly within SharePoint, such as updating or renaming documents when related information changes. For example: you want to rename documents in a folder because a date changed in a document in another folder? Just use the Skills option, it won't disappoint you.
Lesson 3 – Loop, Lists and Teams increase collaboration… and complexity
Ah yes, Microsoft Loop. The quasi-new collaboration app that drives Teams and Copilot and that everybody wants to, or now even needs to use.
In practice, Microsoft Loop is a collaboration tool that surfs on top of SharePoint or OneDrive content. Whenever you make notes via Copilot, or use Facilitator in a meeting, you also make a Loop page on your personal OneDrive or SharePoint environment. Thus Loop makes it easier to co-author, brainstorm, and track work in real time, but it also creates new governance and admin challenges.
Loop is powerful but needs control Administrators can now control who creates Loop content and manage workspaces centrally. However, existing content remains accessible unless specifically restricted.
Loop can fragment information Without clear guidelines, information quickly spreads across Loop pages, Lists, Teams files, and classic SharePoint pages. A single meeting can easily generate multiple Loop components stored in different locations. Sounds inconvenient, right?
Loop can be used in a lot of places, and we really mean a lot Teams chats, channels, meeting, emails, Loop-documents, you name it: you can make a Loop page. This increases the risk of duplication and confusion.
The moral of the story is that without clear rules on where information lives, collaboration tools can reduce findability instead of improving it. This is exactly where our expertise can help you: defining clear content structures on when who uses what, naming conventions, and guidance so that information stays findable across Teams, SharePoint, Lists, and Loop.
Lesson 4 – Discoverability is now a business risk
AI cannot fix poor findability on its own. If content is duplicated, hidden in personal drives, or written unclearly, even the most advanced AI will struggle to deliver trustworthy answers. We've got some practical principles to think about when creating environments for users that will search AI first:
Centralize important content Key procedures, policies, and guides should live in well-governed SharePoint sites, not in personal OneDrive folders or private chats.
Write for search and AI Descriptive titles, meaningful headings, and clear summaries help both people and AI understand the purpose of content.
Keep content fresh and think about a well implemented governance AI-driven SharePoint capabilities can help flag outdated content, but humans still needs to own, review and maintain it overtime.
Putting AI‑ready SharePoint into practice
As Copilot and AI agents become part of everyday work, the real differentiator is no longer the technology itself, but the state of the content you rely on in SharePoint every day. The ECS sessions made this unmistakably clear: Copilot and agents amplify existing content practices, for better or worse. If structure is unclear, ownership is missing, or information is outdated, AI will not compensate for that.
This is why content quality, metadata, ownership, compliance, and automation can’t be addressed in isolation. They need to be tackled step by step, in the right order, if you want AI to retrieve the right information, even when it’s currently buried deep inside documents.
To help you do exactly that, we’ve bundled our expertise and our own workflow into Healthy Docs, Happy AI*: a hands‑on online workshop focused on making your SharePoint library AI-ready.
The workshop consists of four interconnected modules of two hours, running on 6, 8, 13, and 15 October. Each module builds on the previous one and combines clear explanations with concrete examples, exercises, and Q&A, so you can progressively improve the structure, governance, and usability of your SharePoint content as you go.
→ More information about the October workshop session.
*The workshop is based on SharePoint, but the skills you'll learn are applicable to any knowledge base. In-house sessions are also available upon request—please contact us with your needs and context.
Where this is heading, and how Flow supports you
What #ECS2026 made clear is that SharePoint and Microsoft 365 are no longer just collaboration platforms. They are becoming the knowledge layer that Copilot and AI agents depend on. To make the most of this “SharePoint 3.0”, you need the knowledge in your libraries to be accurate, well‑structured, clearly owned, and enriched with the right metadata. That foundation is what allows AI features (Copilot answers to intranet chatbots and automations) to retrieve the right information, instead of amplifying noise.
This shift puts content quality and information design firmly back on the agenda. Not as an abstract ideal, but as a practical requirement for organisations that want AI to be useful, trustworthy, and compliant.
As technical writers and information specialists, we help you strengthen the foundations behind your SharePoint environment by:
Clarifying structure and ownership, so both people and AI know what information to trust
Improving metadata and findability, so Copilot, agents, and chatbots can retrieve the right content
Designing SharePoint and Teams information architectures that work in daily practice and scale with AI
Coaching you in creating and maintaining content that works for people today, and for AI tomorrow
The shift towards “SharePoint 3.0” is less about AI itself and more about the quality of the knowledge it builds on. Organisations that invest in structure, metadata, and ownership today are the ones that will get qualitative, trustworthy value from Copilot and agents tomorrow.
