Psycho-oncology, as an interdisciplinary field, focuses on identifying and managing the psychological, behavioral, and social consequences of cancer and improving the quality of life of patients, and the emergence of generative AI has created new capacities for personalizing psychological interventions in this field. The theoretical basis of using generative AI is based on a person-centered approach and the principle that the cancer experience and the psychological needs of patients are not the same, and factors such as the type and stage of the disease, personality traits, emotional state, coping style, family circumstances, culture, and stage of treatment can create different needs. In this framework, generative AI can help the specialist identify psychological distress, anxiety, depression, fear of relapse, and adjustment problems by processing psychological and clinical data and analyzing the patient's speech and writing patterns. One of the important applications of this technology is to produce psychoeducational content tailored to each patient's circumstances and provide exercises in the field of emotion regulation, stress management, relaxation, problem-solving, and coping skills. Also, AI-based systems can provide the possibility of continuous interaction with the patient and record changes in his/her psychological status in the interval between therapy sessions and report them for expert review. From the perspective of clinical application, this technology can be used as an adjunct tool in early screening, patient education, symptom monitoring, self-management support, and strengthening patient participation in the treatment process. Personalization of interventions means that the content provided is adjusted based on the individual's needs, cognitive capacity, health literacy level, preferences, and mental state, rather than implementing a uniform program for all patients. Early evidence on digital interventions and conversation-based systems suggests that the guided use of AI technologies can be effective in reducing some indicators of anxiety and depression and increasing access to psychological support, but the evidence for generative AI in psycho-oncology is still emerging, and larger studies and long-term follow-ups are needed to draw firm conclusions. One of the key challenges is the possibility of generating misinformation or responses that are disproportionate to the patient's clinical status, which becomes even more important in sensitive topics such as self-harm, mental crisis, cancer prognosis, and treatment decisions. Maintaining the confidentiality of health data, information security, informed consent, algorithmic bias, cultural and linguistic differences, and the lack of clarity of professional liability for errors are also important limitations of this technology. On the other hand, AI-based communication cannot completely replace the human therapeutic relationship, clinical empathy, and expert judgment of the psychologist. Therefore, the use of generative AI should be defined within the framework of human-centered care and as an expert support tool. It is suggested that future systems be designed based on valid psycho-oncology protocols, standard distress assessment tools, and professional ethics principles, and evaluated for accuracy, safety, effectiveness, and reliability before widespread use. It is also necessary to conduct multicentre clinical trials, investigate efficacy in different cancer groups, assess long-term outcomes, and compare AI interventions with usual care. Finally, generative AI can expand the capacity of psycho-oncology by providing flexible, continuous, and tailored psychological support, however, its true value will become apparent when technology is placed alongside psychologist expertise, human communication, and evidence-based clinical decision-making.