May 10, 2026
Barriers and enablers for generative artificial intelligence in clinical psychology: a qualitative study based on the COM-B and theoretical domains framework (TDF) models | BMC Psychology

The COM-B model, in combination with the TDF domains, provided a structure to identify psychologists’ perceptions regarding the adoption of GenAI in care psychology. Eighteen main factors related to the seven TDF domains were identified that significantly influenced the decision to accept or reject the use of GenAI in therapy. Of these, 12 factors belonging to 5 domains functioned as barriers, showing the resistance of some psychologists to integrate GenAI, whereas 6 factors associated with 5 domains functioned as enablers, promoting the acceptance of this technology. Table 4 details the barriers and facilitators identified and classified in the COM-B components, along with the corresponding TDF domains.

Table 4 Significant barriers and enablers identified in the study, classified in the COM-B Components/TDF Domain.

TDF domain: knowledge

The “knowledge” domain refers to psychologists’ understanding of GenAI and its applicability in therapeutic interventions. The data obtained from the interviews indicate that there is a deficiency in this knowledge, which is a significant barrier to the adoption of GenAI. Most psychologists expressed uncertainty or misunderstanding about the capabilities and usefulness of GenAI, affecting their willingness to integrate these technologies into their therapeutic practice. This is illustrated by a participant who expressed concerns about their lack of training:

I am not sure how I could use AI in my practice. I would need more training on what I can and cannot do.” (Interview 1).

This statement reflects a broader trend among participants, emphasizing the need for educational programs to increase awareness and competence in AI tools.

TDF domain: environmental context and resources

The “environmental context and resources” domain highlights external factors that influence psychologists’ willingness to adopt GenAI. One of the most prominent concerns identified was the risk associated with data privacy and security, particularly regarding the confidentiality of session recordings and patient information. As one participant expressed:

“There is a fear of what happens to all information. If a session is being recorded, where is that session going? This information is very confidential and needs to be well explained.” (Interview 1).

This concern underscores the need for clear regulatory frameworks and data protection policies to ensure that AI tools comply with ethical and legal standards in psychological practice.

In contrast, a factor in this domain was identified as facilitator, the potential for AI to reduce administrative workload, allowing psychologists to focus more on patient care. One of the most frequently mentioned benefits was the possibility of having an AI-powered assistant to help with appointment scheduling, documentation, and follow-up tasks. This was illustrated by one participant:

“…it’s that I think it’s super fascinating…, because then there’s also one thing, , writing the medical records is the horror…” (Interview 4).

This statement reflects a broader sentiment among psychologists that administrative duties consume a significant portion of their time, detracting from direct therapeutic work. The use of AI for automating documentation and other routine tasks could be a valuable tool for improving workflow efficiency, reducing burnout, and allowing clinicians to dedicate more time to patient interactions.

TDF domain: Social/professional role and identity

The social/professional role and identity domain focuses on how psychologists perceive their professional roles and identities in relation to the introduction of GenAI. A key barrier identified was the fear of judgments from colleagues, which reflects concern about how other professionals may perceive that using GenAI in therapy may be associated with not being a good professional. One participant expressed concerns about the potential stigma associated with integrating AI into therapy:

“And what judgment would you have if your colleagues used it? that they are not thinking about themselves. I would have that fear…” (Interview 2).

This quote reflects an underlying apprehension about how AI adoption might be interpreted as unprofessional or inconsistent with established therapeutic norms. Such concerns highlight the importance of institutional endorsement and peer discussions to normalize the integration of AI in psychological practice.

The “social/professional role and identity” domain also highlights the importance of institutional endorsement in shaping psychologists’ willingness to adopt AI. A key facilitator identified in the study was the potential role of professional organizations in legitimizing AI-assisted practices and providing ethical guidance. As one participant explained:

“…a good part of my colleagues need a lot of the support of the psychologists’ college or an authority that tells them they can use it and like……” (Interview 6).

This statement underscores the hesitation among professionals to integrate AI into their practice without clear institutional guidance. The endorsement of AI technologies by regulatory bodies or professional associations could provide a sense of security, ensuring that ethical and practical concerns are addressed. Such support may be essential in promoting broader acceptance of AI-assisted interventions among psychologists.

TDF domain: Social influences

The “social influences” domain highlights the role of societal norms and generational differences in shaping attitudes toward AI adoption. One factor that emerged as a facilitator was the perception that younger generations, particularly Gen Z and teenagers, are more open to AI-based interventions. One participant noted this generational shift in attitudes:

“Younger people, more Gen Z and teenagers, that’s why maybe yes, because it truly grinds them the roll and I think they don’t have as much data protection awareness…” (Interview 7).

This statement suggests that digital-native generations may be more willing to engage with AI tools in therapy, as they are accustomed to interacting with technology in various aspects of their lives.

TDF domain: beliefs about capabilities

The “beliefs about capabilities” domain reflects psychologists’ concerns about their ability to adapt to AI-driven changes in their profession. One major barrier identified was the fear of being replaced or professionally devalued as AI becomes more advanced. One participant articulated this concern:

“There’s a feeling, I think, collective that artificial intelligence is doing things that we would never have imagined, and sectors are being removed, right?” (Interview 7).

This statement reflects a broader apprehension about job displacement, a concern that has been observed in other fields integrating AI-based solutions. While AI tools may offer valuable support in administrative and diagnostic tasks, some professionals worry that their expertise could be undervalued or that reliance on AI could alter traditional therapeutic roles. Addressing these concerns through professional development initiatives and clear role delineation could help mitigate resistance to AI adoption in psychological practice.

TDF domain: beliefs about consequences

The “beliefs about consequences” domain highlights psychologists’ concerns regarding the potential negative impact of AI on the therapeutic relationship and patient acceptance. One of the key barriers identified was the perception that certain patient groups, particularly those with paranoia or heightened fears, might strongly resist AI-based interventions. One psychologist illustrated this concern:

“There are patients who wouldn’t want it… most of all, paranoids. I mean, patients with paranoia or with many fears… this wouldn’t be accepted for me.” (Interview 1).

This statement reflects a broader concern among clinicians that AI could be met with resistance from vulnerable patient populations, particularly those with mistrust in technology or heightened anxiety about surveillance. These perceptions suggest that successful AI integration in therapy may require tailored approaches, including transparent communication, informed consent strategies, and careful consideration of patient profiles to ensure ethical and effective implementation.

TDF domain: optimism

The “optimism” domain highlights psychologists’ positive attitudes and curiosity toward the potential of GenAI in therapeutic practice. While some professionals expressed scepticism or concern, others demonstrated a strong interest in understanding how AI could enhance their work. This openness to technological innovation was reflected in the following statement:

“I would see this as extremely interesting… if I saw someone who had incorporated it and used it like that. I’d love to ask you. How do you do it? What do you do? Curious about how they simply, how it is working, and how they react… And how do you use it…” (Interview 4).

This statement suggests that exposure to real-world use cases and peer experiences could play a significant role in fostering AI adoption. Observing how AI is effectively integrated into therapy by colleagues may help reduce scepticism and increase psychologists’ confidence in its application. Future training programs and pilot initiatives could leverage this curiosity by demonstrating concrete examples of AI-assisted interventions in psychological care.

link

Leave a Reply

Your email address will not be published. Required fields are marked *