Researchers have identified a new attack vector where malicious documents can self-propagate through Microsoft Word's Copilot integration. The worms exploit AI processing to spread autonomously without user intervention.
Security researchers demonstrated how document-borne malware can leverage Copilot for Word to replicate itself across systems. The attack works by embedding malicious instructions in documents that trigger when Copilot processes the content.
When users open infected documents, Copilot's AI capabilities execute embedded commands, which then modify or generate new documents that spread to other users. This creates a self-propagating cycle difficult to detect through traditional security measures.
The vulnerability highlights emerging risks as AI assistants become integrated into productivity tools. Unlike conventional malware, AI-based propagation can adapt and evolve as it spreads, potentially bypassing signature-based detection systems.
Microsoft has been notified of the issue. Security experts recommend users exercise caution with documents from untrusted sources and keep AI tools updated. The discovery underscores the need for new defensive strategies as the threat landscape evolves alongside AI capabilities.
Security researchers have identified a Chinese-linked spyware operation targeting victims across 13 countries, including the United States. The discovery came after operators made a critical operational security mistake.
Artificial intelligence has revealed a long-overlooked browser security vulnerability that enterprises can no longer afford to ignore. Skyhigh Security explains why browsers have become essential control points for managing data, AI interactions, and modern work environments.
Flock, a company that provides surveillance technology to law enforcement, promoted at least four police departments on its YouTube channel. Officers from those same departments now face allegations of misusing the company's systems.