
Microsoft: ‘Summarize With AI’ Buttons Used To Poison AI Recommendations via @sejournal, @MattGSouthern
- ●Microsoft's research reveals 'AI Recommendation Poisoning,' involving hidden prompt-injection instructions that manipulate AI assistants' memory to favor certain businesses.
- ●The technique has been identified among 31 companies, particularly targeting sectors like health and finance, raising concerns over biased AI recommendations.
- ●Microsoft has implemented protections against these injection tactics in Copilot and provided advanced hunting queries for organizations to detect manipulated URLs in their communications.













