how to find hidden text in any document.

Each format hides text in its own way, so the fastest check depends on what you were given. This page is the general routine. The step-by-step guides below cover the exact menus and shortcuts for each format.

“Hidden” can describe content that is present in a file but missing from the ordinary reading view. Some cases are intentional. Others are residue from copying, editing, accessibility features, or document conversion.

A finding tells you that the file contains something the ordinary view may not show. It does not, by itself, tell you who put it there or why. Keep that distinction clear before making an accusation or changing a source document.

Word documents are packages of structured content and formatting. What you see on the page is one rendering of that structure. Microsoft provides controls that reveal hidden text and nonprinting marks without requiring you to open the underlying XML.

1. 01Work on a copy
   Duplicate the document first. Preserve the original download so you can compare it or show exactly what you received.
2. 02Show formatting marks
   Use the paragraph-mark Show/Hide control. This reveals spaces, paragraph breaks, tabs, and other nonprinting marks that can explain odd layout or pasted content.
3. 03Display hidden text
   In Word’s display options, enable the setting that shows text formatted as hidden. Microsoft notes that hidden text appears with a dotted underline when displayed.
4. 04Review tracked changes
   Set the review display to show all markup. Inspect comments, insertions, deletions, and formatting changes before accepting or rejecting anything.
5. 05Select suspicious areas
   If a page has unexplained blank space, select across it and inspect the font color, font size, and formatting. Check headers, footers, text boxes, and shapes too.

For the exact menus and shortcuts on Windows, Mac, and Word for the web, seehow to find hidden text in Word.

A PDF often stores positioned text, fonts, graphics, and layers rather than editable paragraphs. Selecting all text and pasting it into a plain-text editor can reveal unexpected reading order or text that was visually covered. Search can also find a word that the page does not appear to show.

This check has limits. A scanned PDF may contain only page images, so there is no text layer to inspect until optical character recognition is applied. Conversely, an OCR layer may be invisible by design because it sits behind the scanned image to make the page searchable. That is normal and should not be treated as suspicious on its own.

1. 01Try selection
   Select the page text and paste into a plain-text editor. Compare the extracted words with the visible page.
2. 02Try search
   Search for distinctive phrases from the visible instructions and move through every result. An off-screen match deserves inspection.
3. 03Check the source type
   Determine whether the file is a digital PDF, a scan with OCR, or a conversion from another document. Each can produce different artifacts.
4. 04Compare a trusted viewer
   Open the file in a current PDF reader. Browser previews and full readers can render layers or annotations differently.

The full walkthrough, with free tools for each check, is inhow to find hidden text in a PDF.

Copying from a webpage or rich document can carry font colors, styles, hyperlinks, comments, and invisible Unicode characters. If a pasted block behaves strangely, paste it as plain text into a new document and compare the result. This removes most visual formatting; it does not prove that every invisible character is unsafe. To see which invisible characters a block carries, run it through Lervan's invisible text checker.

For an assignment prompt, compare three versions: the original file, the ordinary visible text, and a plain-text extraction. Any meaningful sentence that appears in only one version deserves a closer look.

OWASP describes indirect prompt injection as instructions placed in external content—such as a webpage or file—that an AI system later reads. The system may treat those instructions as commands even though the person using it did not write or see them.

That risk is broader than school assignments. A résumé, research paper, support ticket, or shared document can contain text aimed at an automated reader. The practical rule is simple: inspect content before giving it to an AI system, and do not assume visually hidden text is harmless merely because a human reader misses it.

What these instructions usually say, and where they tend to hide, is covered inhow to spot hidden AI prompts.

1. 01Preserve
   Keep the original file, filename, download date, and source location. Take a screenshot that shows the hidden content after it is revealed.
2. 02Classify
   Record whether the finding is hidden formatting, white text, tracked revision content, invisible characters, OCR, or another document artifact.
3. 03Read in context
   Decide whether it changes the visible assignment or is ordinary document metadata or revision residue.
4. 04Ask neutrally
   If the text changes what you are expected to do, ask the document owner which instructions apply. Describe the observation without assuming intent.
5. 05Work visibly
   Use the confirmed visible prompt for your work. Keep drafts and notes that show your process.

Lervan’s free hidden-text checker can inspect Word and PDF files or pasted text for several of these signals. It is a second look, not a verdict about intent, and it cannot guarantee that every kind of visual or structural concealment will be found.

• Save the original prompt and the version you worked from.
• Show formatting marks and review hidden text in Word.
• Resolve comments and tracked changes in your own submission copy.
• Inspect headers, footers, text boxes, and unexplained blank space.
• Export to the required format and reopen the exported file.
• Confirm that citations, page numbers, links, and visible instructions survived conversion.
• Keep drafts or version history until the assignment is settled.

The point is not to become suspicious of every file. It is to know what you are submitting and what an automated reader may extract from it.