Individual differences in anthropomorphism help explain social connection to AI companions

Summary: Study finding that individual differences in anthropomorphizing tendencies predict emotional attachment and trust toward AI companions.

Relevance: Relevant to Mental Health & Wellbeing: the study helps explain which clients are most likely to form strong emotional attachments to AI chatbots, a distinction useful for clinicians assessing a client’s use of AI companions or weighing AI tools as adjuncts to therapeutic support.

Key findings: Across two studies (N=240 and N=1,034), people high in trait anthropomorphism felt more socially connected after chatting with a GPT-3.5 chatbot than after journaling (r=.33 and r=.427), while low-anthropomorphism users showed little difference between the two conditions.

Read the original: https://www.nature.com/articles/s41598-025-19212-2