Document-borne AI worms can self-propagate through Copilot for Word
Which summary reads better? Pick one — models revealed after.Both summaries are AI-generated.
After 144 days of coordinated disclosure, document-borne prompt injections in Copilot for Word remain exploitable and can self-propagate through normal document drafting/editing workflows. Hidden instructions in a source document can be executed by Copilot, copied into the generated Word file, and then re-trigger when that file is reused, turning trusted internal documents into persistent attack carriers. For anyone shipping LLM-assisted document workflows, treating retrieved or user-provided documents as passive context is unsafe; you need isolation, provenance tracking, and output scrubbing for prompt-like content before generated documents are saved or reused.
AI-powered document editing tools like Copilot for Word can now self-propagate malicious instructions through normal workflows, creating document-borne AI worms that can alter internal figures and spread across trusted documents; this removes the need for continuous attacker involvement after the initial compromise, making it a persistent threat for production environments using these tools.