DocumentationExperience Cloud KCS

AEM Forms Output PDF fails to open in Microsoft Edge

Last update: October 5, 2026

In Adobe Experience Manager Forms Output Service, generated PDFs fail to open in Microsoft Edge after network or application-layer modifications corrupt the embedded XMP metadata. The PDFs open correctly in other PDF viewers. To fix this issue, prevent modifications during delivery or remove the XMP metadata as a workaround in specific cases.

Description description

Environment

Adobe Experience Manager Forms Output Service (all versions)

Symptoms

  • PDFs open correctly in most viewers but fail to open in Microsoft Edge.
  • The issue occurs after network delivery, not immediately after PDF generation.

Cause

The PDF is modified in transit, often by HTTP-to-HTTPS rewriting or other text-based response transformations. These modifications corrupt the embedded XMP metadata and invalidate the PDF structure.

Resolution resolution

Follow the steps below to resolve the issue.

  1. Review your application and network delivery path for any response-processing rules that modify PDF content in transit. This includes proxies, web application firewalls (WAF), CDNs, and servlet filters.
  2. Exclude responses with Content-Type: application/pdf (and other binary content types) from any text-based response rewriting or transformations.
  3. Confirm that PDF bytes generated by AEM reach the browser unchanged, without being converted to a string or re-encoded.
  4. Test the PDF in Microsoft Edge to verify that it opens correctly after network modifications are prevented.

If you cannot immediately prevent network modifications, use the following conditional workaround for flat, non-interactive PDFs:

  1. Proceed only if all the following conditions are true:

    • The PDF is flat and non-interactive (not XFA and not dynamic forms).
    • The PDF is not digitally signed, certified, encrypted, or permission-controlled.
    • The PDF is not required for PDF/A or other compliance standards.
    • The application does not depend on XMP metadata for indexing, auditing, routing, archival, or downstream processing.
    • You have completed full regression testing for all forms, templates, and data variations.

  2. Use a PDF processing library (such as Apache PDFBox) to remove the XMP metadata after PDF generation. For example:

    1. Load the generated PDF.
    2. Set the document catalog metadata reference to null (for example,
recommendation-more-help
experience-cloud-kcs-help-kbarticles