The increasing prevalence of mental health disorders has created an urgent need for various modalities of treatment, especially highly scalable, comprehensive, and accessible therapies. Whereas face-to-face methods are still able to provide the best therapy, challenges such as high costs, geographical differences, and shortage of therapists have severely limited accessibility for scalable models. Therapeutics designed with the assistance of conversational agents, passive sensing, or recommendations would enhance effectiveness in figurative terms. The integration of AI into psychotherapy raises concerns about how AI affects the therapeutic relationship, the physician's role, trust, transparency, and privacy. This narrative review will focus on how AI would affect the therapist-client relationship in working alliance and through empathy into trust. While AI will increase efficiency and follow-up on the treatment of patients, AI is limited in giving the in-depth emotional ordinality that humans will confer for therapy to be successful. Furthermore, ethical tensions such as algorithm bias, informed consent, and the propensity to accept AI's recommendations continue to pose other challenges. The results will indicate a hybrid model for AI as a complementary rather than a replacement for professional clinicians. Future studies should take a two-pronged approach by evaluating AI literacy among mental health professionals, making assumption-based arguments for ethical frameworks, and building partnerships between developers and clinicians. A balanced approach will ensure that AI enables therapy while signifying the preservation of core professional values such as trust, empathy, and the human connection in therapy.
Malik et al. (2025) studied this question.