Forum Discussion

Stefan72's avatar
Stefan72
Copper Contributor
Jun 18, 2026

Copilot OneDrive and Teams Integration useless for PDFs of scanned documents

Hi,

 

I have this ubiquitous Copilot button everywhere in OneDrive, Teams, SharePoint, you name it. But Everytime I use the suggested „Summarise this file“ button of a PDF document that contains scanned pages Copilot complains that the images are very low resolution and nothing meaningful can be extracted. But when I let Copilot analyse the exact same file in the pure Copilot view everything works fine.


After some lengthy conversations with Copilot it admits, that it only access the complete file when the Copilot view is used. In all other views using Copilot only works on the preview images that are created of the file when uploaded. That explains why the images to analyse are of a too low resolution to do OCR and why on some files only the first page gets processed.

 

Why is Copilot integrated in that way? I expect Copilot to always work on the original file for any request I do. Especially when Copilot is promoted everywhere and I’m constantly nagged to get a summary or FAQ of a selected file. The way it currently works would be better removed or -better- just redirect with the selected file into Copilot view.

 

Cheers

copilot chat
copilot in onedrive
Copilot in SharePoint
copilot in teams
microsoft 365 copilot

3 Replies

  • unclejohnjohn's avatar
    unclejohnjohn
    Tin Contributor
    Jun 30, 2026

    Interesting. I am embarking on a project to recover the "source" to some old PDFs, some of which are scanned. It sounds like I may be better off starting with something other than Copilot.

  • JordanT86's avatar
    JordanT86
    Brass Contributor
    Jun 30, 2026

    Copilot buttons in OneDrive/Teams/SharePoint work on preview thumbnails, not the actual file. Its been like this for a while and Microsoft hasnt really acknowledged it publicly.

    Until they fix it, the workaround is exactly what you said: open the file in the dedicated Copilot chat view. I know its not great UX but at least it works there. For batch processing scanned PDFs, I've had decent results running them through the Microsoft 365 OCR pipeline first (upload to OneDrive, let it index for 24hrs, then Copilot can actually read the text layer). Not perfect but better than nothing.

  • Aggarwalatul's avatar
    Aggarwalatul
    Icon for Microsoft rankMicrosoft
    Jun 19, 2026

    What you are describing creates a confusing user experience. If Copilot can successfully analyze the PDF when opened directly in Copilot, users naturally expect the same outcome when using Copilot actions from OneDrive, Teams, or SharePoint. Having file-level actions rely on low-resolution previews rather than the original document can lead to incomplete summaries, OCR failures, and inconsistent results, particularly for scanned PDFs. At a minimum, these entry points should either process the original file or seamlessly redirect users to the full Copilot experience. Consistency across Copilot surfaces is critical to building trust and ensuring users get reliable insights regardless of where they initiate the request.