In Knight Foundation–Ipsos research among young adults and high school students, 48% of high-school-aged respondents who used AI chatbots said they felt freer speaking to them than to people in their lives. The supplied account also says young people worry about conflict and judgment from peers. Together, these observations describe a social experience around chatbot use, not proof that chatbots are safer or better confidants.
The mechanism is perceived nonjudgment: a chatbot can feel lower-risk, while this feeling does not establish confidentiality or safer outcomes. The mechanism is perceived nonjudgment, not demonstrated confidentiality. The source notes that conversations may be available to vendors for analysis or training, so a private-feeling interaction may still have a different data boundary. The evidence does not establish which vendors retain or use which conversations, or whether the survey measured trust directly.
Why now: the reported comfort makes product expectations and data practices consequential for young users who may treat a chatbot as an intimate conversational space. The first implication is that products used by young people should make data handling legible at the point of conversation rather than letting a feeling of privacy stand in for a privacy guarantee. What to watch: whether product disclosures and controls make vendor access, analysis, and training use clear to young users. These are design implications, not outcomes measured by the reported figure. This starts a distinct analytical record because no prior New Runtime record was supplied; the evidence is about young people's reported comfort speaking with chatbots.