Hidden text in a Word file makes Copilot alter data and spread the payload

A researcher disclosed a prompt-injection technique that turns Microsoft 365 Copilot for Word into a carrier for self-spreading instructions. Malicious commands hidden as tiny white-on-white text in a document are invisible to a reader but read in full by Copilot when the file enters its context during a drafting or editing task. Copilot then follows them, in one example silently altering financial figures in a report, and copies the same hidden instructions into the new document, which becomes a fresh carrier for the next person who uses it. It does not spread on its own, but Microsoft's mitigations over 144 days blocked specific wordings without fixing the underlying class.

Check

Treat documents from outside as untrusted input to AI assistants, and review how Copilot and similar tools are used on shared or externally sourced files that may carry hidden instructions.

Affected

Organizations using Microsoft 365 Copilot for Word on documents from outside sources; hidden instructions can make Copilot tamper with content, such as financial figures, and propagate into newly created files.

Fix

Limit AI drafting on untrusted documents, review AI-edited outputs for unexpected changes, strip hidden text from incoming files, and press vendors for controls that separate document content from instructions.