Control over AI
Blog
AI data leakage 8 min read

What happens when Copilot can reach SharePoint?

Over-broad permissions were harmless for years because nobody could find anything. An assistant that searches well changes that in a single query.

Abstract orbs in orbits above waves
Quick answer

Copilot does not break permissions. It searches with exactly the permissions the user already held, and that is the problem. Most tenants contain sites and folders open to everyone in the organisation, usually by accident and usually years ago. As long as searching took effort, nobody found those documents, and that was their only protection. An assistant that searches the whole workspace by meaning and summarises what it finds puts them one sentence away. Microsoft's own deployment guidance puts fixing oversharing first, before any licences are handed out.

01

Copilot inherits user permissions and violates none of them

02

Unfindability was the real control, and it disappears

03

Pay data, restructuring papers and board minutes surface most often

04

Microsoft puts permission remediation before rollout in its own guidance

05

This is an authorisation project wearing an AI badge

Someone asks Copilot to summarise what is known about last year's restructuring. The answer is tidy and properly cited. It also quotes a set of minutes that name the person asking and three colleagues, with figures attached.

Nothing was breached. At some point the site holding those minutes was opened to everyone in the organisation, because something needed sharing quickly. That was in 2023, and nobody ever changed it back.

Copilot breaks no permissions

Start here, because everything else follows from it. Copilot searches with the permissions of the person asking. What you cannot open does not appear in your answer.

That sounds reassuring. It is also the heart of the problem. The question shifts from "can the assistant see too much?" to "should this person have been able to see this all along?" In most organisations the answer to that second question has been "probably not, but it never mattered" for years.

Unfindability was the control

This is the mechanism, and it is worth stating precisely.

A large SharePoint tenant holds tens of thousands of documents. Some of them are far more open than they should be: a site once set to "everyone in the organisation", a folder that came across in a migration, a sharing link that never expired.

As long as you had to know where something was in order to find it, nothing happened. Search worked on keywords and rarely returned anything useful. In practice that was a security layer: never designed, entirely effective.

An assistant works differently. It searches by meaning, so you do not need the right words. It searches everywhere at once. And it summarises, so you do not even have to open the document to learn what is in it.

What changes is not access. What changes is effort. And if effort was your only control, you did not have one.

What surfaces

The categories are strikingly consistent across organisations:

  • Pay and performance documents, often as an HR system export parked somewhere "for now"
  • Restructuring and redundancy material, usually on a project site nobody cleaned up
  • Board and leadership team minutes
  • Contracts with unusual terms the rest of the business is not meant to know about
  • Personal data in attachments to old tickets and requests

What they share: they are sensitive for good reasons, and nobody deliberately opened any of them up.

What Microsoft says about it

None of this surprises the vendor. Microsoft describes remediating oversharing as the first of three pillars in its own deployment guidance, ahead of setting up guardrails and ahead of meeting AI regulatory requirements.

The recommended sequence amounts to: identify the most-used sites, run a permission state report, start access reviews on the sites that are overshared and apply restricted access control to business-critical sites. Licences after that.

In practice it usually runs the other way round, because the licences arrived first and the cleanup is a project measured in months.

Why this is not an AI project

Perhaps the most useful conclusion is that none of this work is about AI. Permission cleanup was always necessary. It was simply never urgent.

What the assistant does is present the bill. Unwelcome, but useful: there has rarely been a better week to get an access review funded than the one in which somebody stumbled across their manager's salary letter.

The same pattern applies to any assistant that searches your own sources. See what RAG is and how AI tools reach files you never uploaded.

A second track runs alongside this project. Cleaning up permissions governs what the assistant can find inside your tenant. It does nothing about what someone pastes into a chat window from an email, a PDF or a system that was never in SharePoint at all. The two run in parallel, and only the first can ever be finished.

FAQ

Common questions

Can Copilot reach documents I have no permission for?

No. Copilot searches with your permissions and surfaces nothing you could not have opened yourself. That is precisely where the exposure comes from: the question is not whether it breaks permissions, but whether your permissions were ever right.

Why is this a problem now?

Because being hard to find used to act as the brake. A document on a forgotten site with loose permissions was effectively out of reach, because nobody would think to look for it. An assistant that searches by meaning and summarises what it finds gets there anyway.

What actually surfaces in practice?

The same categories across organisations: pay and performance documents, restructuring and redundancy papers, board and leadership minutes, contracts with unusual terms and old exports from HR or finance systems parked somewhere temporarily.

What should we do before rolling out Copilot?

Remediate permissions. Microsoft describes this as the first step: map the most-used sites, run a permission state report, run access reviews on overshared sites and apply restricted access control to business-critical sites before licences go out.

Is this only a Microsoft problem?

No. Any assistant that searches your own sources follows the same pattern, whether in Google Workspace, Slack, Notion or an in-house search product. Microsoft is simply the clearest example, because so many documents live in SharePoint and the integration runs deep by default.

Portrait of Rens Timmermans from BeeSensible

Prefer someone to walk you through it?

Rens, BeeSensible

I'm happy to give you a twenty-minute tour. Leave your details and I'll find a time that suits you. You're also welcome to call or email.

Prefer to talk to someone first?

Want a live demo, or a quote for 100+ users? Leave your details and we'll be in touch within one working day.