Study Reveals Impact of Attachment Styles on AI Counseling Tool Adoption

Attachment styles shape adoption of AI counseling tools, study finds

Attachment styles and level of trust in AI influence how willing people are to use it as a counseling tool, a new UC study has found. Led by Te Whare Wānanga o Waitaha, University of Canterbury (UC) MSc Psychology student Xiaoli Wu and her supervisors Dr. Kongmeng Liew and Dr. Martin Dorahy from the School of Psychology, Speech and Hearing, their research has shown that people with particular attachment styles are more likely to adopt artificial intelligence (AI) tools for support. The findings are published on the preprint server JMIR AI. Conversational AI, powered by large language models, is already being used by many as a therapy tool-whether it has been built for this or not. It provides accessible, cost-effective, and fast alternatives to traditional therapy, addressing barriers such as stigma and lack of resources. Wu wanted to understand the varied uptake of AI for counseling and explore psychological factors driving adoption. "We surveyed 239 American adults who had not previously used CAI for mental health support, and assessed their trust in AI, their attachment styles -known as anxious or avoidant, and willingness to adopt AI counseling," says Wu. "Higher trust in CAI significantly increased the likelihood of adoption, if a participant trusted the AI's competence and safety, they were more likely to accept it for mental health support. Anxious attachment, characterized by heightened preoccupation with relationship security and fear of abandonment, was positively associated with the use of CAI. Whereas avoidant attachment -- marked by discomfort with closeness, suppression of attachment needs, and reluctance to depend on others, despite underlying concerns about relationships -- showed no significant effect," says Dr. Liew. "We think that individuals with anxious attachment may view CAI as a safe, non-judgmental alternative to human therapists. "While these findings are promising for helping with the overloaded mental health system, our research still displayed gaps, including data privacy concerns and the potential for emotional over-reliance, and the need for rigorous research to enhance the safety, reliability, and personalization of AI-driven counseling services," says Wu. "This study underscores the significance of psychological factors, like trust and attachment styles, in shaping attitudes toward CAI. Understanding these elements can guide the development of more effective and user-friendly mental health technologies."

Attachment Styles Shape Adoption Of AI Counselling Tools

Led by Te Whare Wānanga o Waitaha | University of Canterbury (UC) MSc Psychology student Xiaoli Wu and her supervisors Dr Kongmeng Liew and Dr Martin Dorahy from the School of Psychology, Speech and Hearing, their research has shown that people with particular attachment styles are more likely to adopt artificial intelligence (AI) tools for support. Conversational AI, powered by large language models, is already being used my many as a therapy tool-whether it has been built for this or not. It provides accessible, cost-effective, and fast alternatives to traditional therapy, addressing barriers such as stigma and lack of resources. Wu wanted to understand the varied uptake of AI for counselling and explore psychological factors driving adoption. "We surveyed 239 American adults, who had not previously used CAI for mental health support, assessed their trust in AI, their attachment styles -known as anxious or avoidant, and willingness to adopt AI counselling," says Wu. "Higher trust in CAI significantly increased the likelihood of adoption, if a participant trusted the AI's competence and safety, they were more likely to accept it for mental health support. Anxious attachment, characterised by heightened preoccupation with relationship security and fear of abandonment, was positively associated with the use of CAI. Whereas avoidant attachment -- marked by discomfort with closeness, suppression of attachment needs, and reluctance to depend on others, despite underlying concerns about relationships -- showed no significant effect." says Dr Liew "We think that individuals with anxious attachment may view CAI as a safe, non-judgmental alternative to human therapists. "While these findings are promising for helping with the overloaded mental health system, our research still displayed gaps, including data privacy concerns and the potential for emotional over-reliance, and the need for rigorous research to enhance the safety, reliability, and personalisation of AI-driven counselling services," says Wu. "This study underscores the significance of psychological factors, like trust and attachment styles, in shaping attitudes toward CAI. Understanding these elements can guide the development of more effective and user-friendly mental health technologies." To read the paper click here: https://preprints.jmir.org/preprint/68960/accepted?_cldee=CU3EEVRxubrV4HHJhgBcfqlRfs4nXHLDM6e7MIkVcqILjKaLWgd676JJQTbgMlk5&recipientid=contact-af731e301a68ea11a811000d3ad199c8-55568a7019374acdb9004ce834759857&esid=e3934f68-3313-f011-998a-6045bde553dd