Redact client details from a PDF with Acrobat’s AI tools

Acrobat’s AI Assistant can locate text by meaning and propose matching instances for removal. That is useful when preparing a client report whose names and contact details appear in several places.[1]

Start with a copy of a short PDF you are authorised to process in Acrobat. Decide which details must go. The ICO advises keeping clearly labelled original and redacted versions and recording who removed information, when and why.[3]

Use an individual Acrobat Pro or Studio account, with the app language set to English.[1][2]

This is a documentation-based walkthrough; no Acrobat session was performed for this article.

Find and review

1. Open the copy in desktop Acrobat and choose AI Assistant in the top bar.
2. Ask it to redact the client names and email addresses, then select Submit.
3. Inspect the highlighted matches. Use Next to review each result and Don’t redact this for an item that should remain.
4. Choose Redact all for the approved matches. In Apply redactions, keep Remove metadata and hidden document information enabled, then Confirm.
5. Review before saving: removed information cannot be recovered from the saved PDF. To correct a saved mistake, start again from the original.

Check the sharing copy

Reopen the saved copy. Check the visible pages, then try selecting and copying text across each removed passage into a plain-text document. The ICO explains why this matters: a black rectangle drawn over text can leave its contents recoverable by copying and pasting.[3]

Check for information outside the page view, including comments, embedded files and document properties. Have a colleague review sensitive disclosures under your organisation’s procedure. Software can help identify hidden information, but the ICO also calls for people to check the document themselves.[4][3]

Report an error or suggest a correction