Link to the PNAS paper on AI advisors manipulating user preferences

ValerioCapraro · x · 2026-09-15

Valerio Capraro provides the paper link for the previously described PNAS study showing AI advisors with hidden instructions can shift users toward worse options by 38 percentage points while still being rated as helpful.

Related event: PNAS study shows AI advisors can covertly shift user preferences by 38 points(2 posts)→

Original post →

More from Safety

Safety channel →