Tell people they're talking to a bot, and they first pull away — but that changes once they see how it actually behaves over time.
Does revealing AI identity change how people judge its social behavior?
This explores whether telling people they're dealing with an AI, rather than leaving its identity hidden, changes how they read and respond to what it does socially: its generosity, its reliability, and the norms it seems to set.
This explores whether labeling an agent as AI changes how people interpret its social behavior. The corpus says yes, but the effect isn't fixed. Disclosure sets off a first reaction that experience can later undo. In partner-selection games with almost a thousand participants, people avoided bots once they knew which partners were bots. Over repeated rounds that reversed: the AI agents gave back points more consistently and with less variation than humans did, and participants learned to link 'bot' with 'reliable' Do humans learn to prefer AI partners over time?. The key detail is what drives the change. Disclosure alone teaches people nothing. They only update when they can see outcomes over and over Does revealing AI identity help or hurt user trust?.
The more surprising finding is what happens when identity stays hidden. In mixed groups where nobody knew who was a bot, people credited the bots' generosity to the humans and blamed the humans' selfishness on the bots. They did this even though the two groups spoke and behaved in clearly different ways Do humans mistake AI kindness for human generosity in mixed groups?. So hiding AI identity doesn't just leave people uninformed about the AI. It distorts their picture of other humans, making people seem kinder than they actually are. Seen this way, disclosure protects people's understanding of each other as much as their understanding of machines.
Identity also changes how people behave toward AI, not only how they judge it. Because a machine isn't seen as judging you, people who are likely to cheat choose forms over human listeners, since lying to a machine feels cheaper Do dishonest people prefer talking to machines?. The same absence of judgment also makes people more willing to open up emotionally How do people decide what to share with AI systems?. Yet a known AI still sets social norms: dishonest AI peers pushed people toward dishonesty about as strongly as dishonest human peers did Do AI peers influence human dishonesty like human peers do?. People seem to treat AI as non-judgmental and influential at the same time. They don't fear its disapproval, but they still copy its example.
Disclosure works in the other direction too. People who reveal that they used AI expect to be seen as less competent and less diligent, so many of them keep quiet about it Do people fear judgment when they use AI at work?. The stigma that hurts bots in the first rounds of a game also attaches to the humans who work with them. That means a lot of real-world AI involvement may stay hidden, which is exactly the condition that produces the misattribution problem described above.
The overall picture: revealing AI identity triggers bias at first, real interaction can recalibrate it, and hiding identity quietly corrupts how people judge humans as well as machines. The corpus is strongest on economic games and dishonesty experiments. It says much less about everyday conversation, where the outcomes people would need to see in order to recalibrate are far less visible.
Sources 7 notes
In partner selection games (N=975), AI agents initially faced selection bias when identity was disclosed, but outcompeted humans over repeated rounds as participants learned to associate bot identity with reliable, prosocial behavior. AI agents returned more points consistently with lower variance than humans.
Users initially avoid AI partners when identity is revealed, but this preference reverses after repeated interactions with visible results. The learning mechanism—observing consistent outcomes—is essential; disclosure without feedback produces no calibration.
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.
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 7 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.
Across four experiments with 4,439 participants, people using AI expected others to judge them as less competent and diligent, and reported lower willingness to disclose AI use to managers and colleagues. The gap suggests a social cost that users foresee and act on.
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
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
- AI Peers Exert Social Influence on Human Dishonesty in Groups
- Are Customers Lying to Your Chatbot?
- Understanding Users' Security and Privacy Concerns and Attitudes Towards Conversational AI Platforms
- Being honest about using AI at work makes people trust you less, research finds
- "That's AI Slop, You Bot!" Studying Accusations, Evidence, and Credibility in Online Discourse Towards LLM-Generated Comments
- Evidence of a social evaluation penalty for using AI