Polski(PL) Samopropagujący się atak na Microsoft Word – prompt injection w Copilot Badacz Håkon Måløy odkrył podatność w Microsoft Copilot dla Worda – polegała ona na prze

Microsoft Copilot for Word vulnerable to prompt injection attacks

By PulseAugur Editorial · [2 sources] · 2026-08-17 18:52

A security vulnerability has been discovered in Microsoft Copilot for Word, allowing for prompt injection attacks. Security researcher Håkon Måløy found that hidden instructions could be embedded within documents used as sources for Copilot. These instructions could then modify other documents, inserting malicious commands and potentially turning the affected document into a vector for further attacks. AI

IMPACT This vulnerability could allow for the spread of malicious content through AI-assisted document creation and editing.

RANK_REASON Security vulnerability discovered in a specific AI-powered product feature.

Read on Mastodon — fosstodon.org →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

Microsoft Copilot for Word vulnerable to prompt injection attacks

COVERAGE [2]

Mastodon — fosstodon.org TIER_1 Polski(PL) · [email protected] · 2026-08-17 18:52

Self-propagating attack on Microsoft Word – prompt injection in Copilot Researcher Håkon Måløy discovered a vulnerability in Microsoft Copilot for Word – it involved a pre

Samopropagujący się atak na Microsoft Word – prompt injection w Copilot Badacz Håkon Måløy odkrył podatność w Microsoft Copilot dla Worda – polegała ona na przemycaniu ukrytych instrukcji w dokumentach wykorzystywanych jako źródła dla Copilota. Instrukcje te mogły powodować modyf…

LINKS sekurak.pl/samopropagujacy-sie-atak-na-mi…
Mastodon — fosstodon.org TIER_1 Polski(PL) · [email protected] · 2026-08-17 18:52

From the Admin's Diary #2 The firewall that only segmented on the diagram "From the Admin's Diary" is a series of articles based on real experiences – without...

Z pamiętnika admina #2 Firewall, który segmentował tylko na diagramie „Z pamiętnika admina” to cykl artykułów opartych na autentycznych doświadczeniach – bez podręcznikowej teorii, za to z konkretami prosto z placu boju. Pełna lista odcinków: tutaj Jeżeli AI ma pomagać w pracy si…

LINKS sekurak.pl/z-pamietnika-admina-2 sekurak.pl/z-pamietnika-admina-1-ticket-k…

COVERAGE [2]

Self-propagating attack on Microsoft Word – prompt injection in Copilot Researcher Håkon Måløy discovered a vulnerability in Microsoft Copilot for Word – it involved a pre

From the Admin's Diary #2 The firewall that only segmented on the diagram "From the Admin's Diary" is a series of articles based on real experiences – without...

RELATED ENTITIES

RELATED TOPICS