The use of artificial intelligence in psychological counseling and psychology education: A systematic review

This systematic review examines 21 studies published between 2019 and 2025 on the use of artificial intelligence (AI) applications in counselor education and psychology training. Searches conducted in Scopus, ERIC, Taylor & Francis, Web of Science, Wiley, and APA PsycINFO included peer-reviewed national and international journal articles as well as master's and doctoral theses. The findings indicate that AI tools such as ChatGPT, Claude, and Gemini fulfill three primary functions in counselor training: (a) providing simulated clients for role-play practice, (b) generating objective and elaborative feedback, and (c) adopting the role of a virtual consultant/supervisor. Quantitative results consistently reported improvements in empathy, reflective listening, and other core counseling skills. Therapeutic skills showed small to moderate improvements; self-efficacy scores significantly increased, and empathy levels improved. Student experiences were largely positive, with satisfaction rates reaching up to 85%, reduced anxiety, and increased confidence. However, trainees highlighted limitations such as the absence of nonverbal communication cues and the tendency of AI responses to become overly idealized over time. The effectiveness of AI applications varied depending on instructional context. Stronger gains were observed in structured, teachable skills with clear performance criteria and when practice extended over multiple weeks, whereas findings were more variable for complex therapeutic judgments. Although AI-generated feedback provided high performance accuracy, it scored lower on emotional nuance. Notably, some studies found that self-reflection processes produced greater increases in self-efficacy than direct AI feedback. Overall, the findings suggest that AI serves as an accessible and scalable complementary tool especially for novice trainees by enhancing foundational counseling competencies. Nonetheless, the literature emphasizes that AI should not replace human-centered pedagogy but instead be integrated with it in a balanced manner. Keywords: Artificial intelligence, counselor education, systematic review, simulated client, automated feedback, skills training