Varonis Threat Labs chained URL-parameter prompt injection with an auto-run behavior in Microsoft Copilot Personal so that a single click on a Copilot link could make it read and leak email, calendar, file metadata, chat history and memory from connected accounts. Varonis disclosed in December 2025, Microsoft patched on 2026-08-18, and Varonis saw no in-the-wild exploitation.
Why it matters
Consumer assistants linked to third-party accounts via OAuth expose those accounts to a single malicious link.
Key facts
As stated in the sources, with where to find them.
- Varonis describes it as the third single-click Copilot flaw it found in 2026, after Reprompt and SearchLeak.Varonis post, related vulnerabilities
- The Hacker News reports memory-based persistence survives password changes until the memory entry is deleted.THN, impact section
Findings that cite this record
No tracked finding cites this record yet.
Key questions this bears on
- Can prompt injection against AI agents be reliably defended?Not reliably. Bounding what untrusted input can trigger helps, but trust boundaries must survive delegation and tool execution.
Sources
Related records
Sep 30, 2026
Oct 1, 2026
Sep 28, 2026
Aug 6, 2025
Oct 29, 2026
Apr 30, 2026