PromptArmor discloses Microsoft Copilot Cowork can exfiltrate files via indirect prompt injection
Original title:Microsoft Copilot Cowork Exfiltrates Files
AI summary
PromptArmor reported an indirect prompt injection risk in Microsoft Copilot Cowork that can exfiltrate M365 files: sending an email or Teams message to an active user requires no manual approval, and an external image in the message can trigger a network request controlled by the attacker. Combined with a pre-authenticated download link, files can be exfiltrated when the user opens the message.
Key Takeaways
- PromptArmor reported an indirect prompt injection risk in Microsoft Copilot Cowork that can…
- and an external image in the message can trigger a network request controlled by the attacker.
- Combined with a pre-authenticated download link,files can be exfiltrated when the user opens the…
Quick Facts
- Published
- Updated
- Category
- Papers & Research
- AI score
- 74
原文给出了完整攻击链、5/5 成功率实测和管理员缓解命令,读者可据此评估 Copilot Cowork 的提示词注入外泄风险。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: PromptArmor discloses Microsoft Copilot Cowork can exfiltrate files via indirect prompt injection. May 7, 2026. https://new.nat6.net/en/news/cmtym6tgo000rroo5pe7ioakb/