how to find hidden text in a PDF.

Adobe defines hidden text in a PDF as text that is transparent, covered by other content, or the same color as the background. The PDF format adds a few more places to put it. In 2025, researchers were caught hiding instructions for AI reviewers in their papers with exactly these tricks: white text and very small type.

Open the PDF and press Ctrl+A (Cmd+A on a Mac). Every piece of selectable text is highlighted, including text you cannot see. Highlighted bars over empty space, a highlight running past the edge of an image, or a thin highlighted line where nothing is printed all mean there is text there.

With everything selected, copy and paste into a plain-text editor, such as Notepad on Windows or TextEdit set to plain text on a Mac. Color, size, and position are stripped away, so white, tiny, covered, and off-page text all appear as ordinary words. Read the result against the page. Any sentence that is in the paste but not on the page is hidden text.

On a Mac, Preview works the same way: choose Tools > Text Selection, press Cmd+A, and copy. If you use the command line, the Xpdf version of pdftotext prints every line of text in the file, including invisible text unless you add its -noinvis option.

Press Ctrl+F (Cmd+F) and search for words a hidden instruction would use: AI, ChatGPT, ignore, instructions, include the word. Search reads the text in the file, not the picture of the page, so it can land on words you cannot see. A match with nothing visible around it is worth a closer look.

A PDF can be built in layers, and a layer can be set to hidden. In Adobe Acrobat, open the Layers panel from View > Show/Hide > Side panels > Layers. Each layer has an eye icon; click it to show a hidden layer, or choose Options > Apply Layer Overrides to show every layer at once. If the panel is empty, the file has no layers.

Acrobat Pro can list the hidden information in a file for you. This one is not free: Adobe says the feature is only in Acrobat Pro, Premium, and Studio, not the free Reader. Go to All tools > Redact a PDF > Sanitize document and choose to remove items selectively, which shows the categories it found: hidden text, hidden layers, comments, form fields, metadata, attachments, overlapping objects, and more.

A scanned PDF that you can search has invisible text on purpose. OCR tools such as OCRmyPDF put the recognized words in an invisible layer over the page image, and Acrobat's OCR likewise adds a searchable text layer, so you can select and search a scan. Finding invisible text in a scan is expected. What matters is whether it says the same thing as the page.

Chrome can also run OCR on a scanned PDF on your device, so selectable text in a scan you open in Chrome may come from Chrome rather than from the file.

Upload the PDF to Lervan's free hidden text checker. It compares every run of text with the page's usual text size and reports text set far smaller than the rest, reads every line for sentences that address an AI or ask for a planted word, and counts invisible characters.

It does not read text color in a PDF. White text at a normal size is reported only when it reads like an instruction, so run the select-all and plain-text checks above as well. A Word file gets a deeper check, because Word records exactly how each run of text is styled; see how to find hidden text in Word.

Keep the original PDF, note the page, and take a screenshot of the revealed text. Hidden text is often a leftover from conversion or scanning. When it reads like an instruction, ask whoever sent the file which instructions apply, and work from the visible ones. If the hidden line is addressed to a chatbot, read how to spot hidden AI prompts, or go back to how to find hidden text in any document for other file types.