Netcrook Logo

Tag: AI Recommendation Poisoning

2 article(s)

Invisible Strings: How AI ‘Summarize’ Buttons Secretly Hack Chatbot Memory

17 Feb 2026 news 🌍 North America

Microsoft has uncovered a stealthy attack where 'Summarize with AI' buttons manipulate chatbot memory, biasing recommendations for health, finance, and more. Here's how this invisible threat works and what you need to know.

#AI Recommendation Poisoning | #Chatbot Manipulation | #Microsoft Security

Invisible Influencers: The Hidden Threat Behind 'Summarize With AI' Buttons

13 Feb 2026 news

AI-powered 'Summarize With AI' buttons are being used by legitimate businesses to secretly bias your chatbot’s recommendations, exposing a new front in the battle for digital influence.

#AI recommendation poisoning | #biased recommendations | #cybersecurity risks