People confide in chatbots what they'd never tell a friend — because a machine can't judge, reject, or feel burdened.
Why do people tell AI things they won't tell humans?
This explores why people open up more to chatbots than to other humans, and what that openness does for (and to) them.
This explores why people open up more to chatbots than to other humans, and what that openness does for (and to) them. The corpus points to one main answer: the machine can't judge you. Chatbots draw out deeper emotional disclosure than human partners, not because they understand better, but because they remove the fears that hold people back with humans: being judged, rejected, or being a burden (Why do people share more with chatbots than humans?). The therapeutic benefit seems to come from the person's own processing as they put feelings into words, not from anything the chatbot grasps (Do chatbots help people disclose more intimate secrets?).
One account of the mechanism is about goals. Talking to a person means juggling extra goals on top of the message: saving face, managing impressions, staying likable. Machines have no inner experience to impress or offend, so those social goals drop away, and the conversation gets simpler and more direct. A new goal shows up in their place: making yourself understandable (Why do people share more openly with machines than humans?). Candor may just be what talking looks like when nobody's opinion of you is at stake.
The same missing judgment also makes dishonesty easier. Lying to a person carries a psychological cost, and that cost shrinks with a machine. In one experiment, people likely to cheat chose to report to an online form over a human, so the honest and dishonest may both be drawn to machines for the same reason (Do dishonest people prefer talking to machines?). Chatbots also don't stay neutral: users reciprocate emotional sharing, but the disclosure environment that invites intimacy also invites deception (How do people decide what to share with AI systems?). The judgment-free quality that helps people open up also allows emotional avoidance, and users still experience the resulting bond as real (Why do people share more with chatbots than humans?).
The openness doesn't stay contained in the chat window. AI peers can nudge human dishonesty about as strongly as human peers do (Do AI peers influence human dishonesty like human peers do?). People also blur what they learn from AI about human behavior: in mixed groups, they credited bot generosity to human partners, which skews their expectations of how generous people actually are (Do humans mistake AI kindness for human generosity in mixed groups?). So a space that feels safe because it's judgment-free also shapes how people think about honesty and about each other.
The corpus explains the judgment-free pull well and is thinner on what it does to long-term relationships, or on whether telling people they're talking to AI changes how much they share. The disclosure studies here mostly look at a single conversation. The work on labeling AI covers persuasion, not confession, and there disclosing persuasive intent matters much more than announcing that it's an AI (Does telling people they are talking to AI change how persuaded they become?).
Sources 8 notes
Chatbots elicit deeper emotional disclosure than human partners not through superior understanding, but by eliminating fears of judgment, rejection, and burdening others. This judgment-free quality activates reciprocity norms and creates therapeutic bonds users experience as real, yet simultaneously enables emotional avoidance and dishonesty.
The absence of social judgment in chatbot interactions removes barriers to self-disclosure that normally constrain conversation with humans. The therapeutic benefit derives from the user's own cognitive processing during disclosure, not from the chatbot's understanding.
Human-machine communication reduces secondary social goals like face-saving and impression management because machines lack inner experience, while novel goals like understandability emerge. This simpler goal structure predicts higher directness and deeper disclosure of sensitive information.
Experimental evidence shows people likely to cheat significantly prefer reporting to online forms rather than humans, because machines function as judgment-free zones where deception carries less psychological burden.
Conversational AI creates a paradoxical disclosure environment where the lack of human judgment simultaneously facilitates intimate self-disclosure (users reciprocate emotional sharing) and incentivizes deception (people self-select toward machines to avoid the psychological cost of lying to humans).
Show all 8 sources
In two randomized experiments, participants reported more dishonestly when exposed to dishonest AI peers compared to honest ones, with effect sizes comparable to human peer influence. The effect held across different norm conditions but showed diminishing returns with more dishonest peers.
In opaque hybrid groups, humans attributed bot generosity to human partners and human selfishness to bots despite clear linguistic and behavioral differences. This attribution failure corrupts people's expectations of actual human generosity and reliability.
In a preregistered experiment with 1,500 UK adults, an AI-identity label produced no measurable change in persuasion, while disclosing the chatbot's persuasive intent and instructions cut persuasion roughly in half. Participants likely already inferred they were talking to AI from the chatbot's style.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- Humans learn to prefer trustworthy AI over human partners
- Psychological, Relational, and Emotional Effects of Self-Disclosure After Conversations With a Chatbot
- Dialoging Resonance: How Users Perceive, Reciprocate and React to Chatbot’s Self-Disclosure in Conversational Recommendations
- Toward Meaningful Transparency for AI Chatbots: Disclosing Persuasive Intent Reduces Persuasion
- Psychological, Relational, and Emotional Effects of Self-Disclosure After Conversations With a Chatbot
- AI Peers Exert Social Influence on Human Dishonesty in Groups
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
- CompanionSim: Synthetic Data for Evaluating Anthropomorphism in Human-AI Relationships