Ethical challenges in using neural networks in the work of a psychologist

Neural networks open new opportunities for psychologists, aiding in session analysis, material preparation, education, and service promotion. The use of artificial intelligence (AI) in psychology, particularly in transactional analysis (TA), is associated with ethical risks: data breaches, algorithmic bias, lack of transparency in decision-making, and diminished trust from clients and therapists. This paper explores how neural networks can assist in TA practice while maintaining therapeutic presence and adhering to ethical principles. Special attention is given to confidentiality, responsibility, bias, and legal issues. The risks related to AI’s lack of empathy and its potential impact on trust in the therapeutic alliance are examined. Recommendations for the safe use of AI include data anonymization, employing local systems, supervision, and limiting AI to routine tasks. The importance of preserving the human factor in psychotherapy, where empathy, intuition, and professionalism remain irreplaceable, is emphasized. The study highlights the need to comply with SOTA and EATA ethical codes and data protection legislation. AI is viewed as a tool that enhances practice but does not replace the therapist, underscoring the balance between automation and human interaction in TA.