Affiliation in human-AI interactions based on shared psychological traits

People affiliate with others who share their psychological traits. Does the same phenomenon occur with AI instructed to mimic human psychology? Large language models (LLM) were prompted to use language that mimicked anxious symptoms or their absence (Experiment 1; n=100), extroversion or introversion (Experiment 2; n=100), and an exact mirror or inverse of each participants’ personality (preregistered Experiment 3; n=100). In each experiment, participants engaged in online written interaction with both versions and then evaluated their engagement with the AI. Those with anxiety reported a stronger connection to the LLM that mimicked anxiety, a distinction also reflected in the sentiment of the messages they exchanged. Extroverted participants affiliated more with the AI that mimicked extroversion. Finally, when participants interacted with LLMs that mimicked either their own Big Five personality profile or the opposite of their personality, they reported more affiliation with their doppelganger; this distinction was reflected in the sentiment of their messages. Results support affiliation in human-AI interactions based on the linguistic presentation of a shared psychology. We propose that through socioaffective tuning, LLMs might achieve greater human-like correspondence.