
.webp)


.webp)

Consumers show highest confidence in AI when it assists rather than decides. Writing and summarizing leads at 37% comfort, followed by search at 32% and shopping recommendations at 24%.
The pattern is clear: AI works best as an assistant, not an authority. Users trust AI to help them think, organize, and explore—but expect humans to review before high-stakes decisions.
Privacy emerges as the leading concern (33%), followed closely by inaccuracy and made-up information (24%), and job displacement (22%). These three issues account for 79% of recorded AI distrust.
Notably, distrust varies by age and education. Gen X worries most about accuracy; Millennials about job replacement; higher-education consumers about privacy. Trust is not universal—it shifts by demographic.
The consensus is stark: when the stakes rise, humans expect to stay in control. 33% identify medical diagnosis as where AI should never decide alone. 31% point to loans and mortgages. 32% name hiring and firing.
40% of users prefer AI to make recommendations only—leaving the final decision to a human. Another 30% would allow independent AI decisions only in low-consequence areas. The message is clear: capability does not equal permission.