The integration of artificial intelligence (AI) into psychotherapy holds promise for revolutionizing mental healthcare delivery, particularly in addressing the pervasive treatment gap and improving accessibility to services. This paper critically examines the landscape of AI utilization within psychotherapy, focusing on its potential benefits, ethical dilemmas, and optimal conditions for integration. AI systems offer valuable support in diagnosis, treatment planning, and intervention personalization, yet navigating the boundaries between AI-driven interventions and human-centered therapy requires careful consideration. Factors influencing AI implementation include the nature of mental health issues, patient attitudes towards technology, and environmental factors such as stigma and therapist shortages. We discuss potential pitfalls, including ethical concerns, power dynamics, and inherent biases in AI algorithms, and underscore the  importance of maintaining the therapeutic alliance. Moreover, we highlight the synergies between AI-based and human therapy, emphasizing the need for a balanced approach that prioritizes patient welfare and preserves interpersonal connection. 

The development of large language models (LLMs) has led to the proliferation of chatbot services like ChatGPT, Replika and Project December further contributing to technologically mediated grief. Called variously griefbots, or thanabots, the postulated aim behind these technologies is to help people deal with the loss of their loved ones. Despite hype around these technologies in the media, little is still known about the actual mechanisms of how thanabots contribute to grieving. This paper contributes to answering this question by offering an interpretation of communication with thanabots through the prism of Yuri Lotman's autocommunication model. We start by examining how a thanabot can be framed as a communicational `other'. Then we draw on phenomenology of grief literature to delineate the boundaries of such ``otherness''. In line with the Lotman's auto-communication model, we proceed to argue that conversing with a thanabot can be viewed as an instance of `I--I' communication. Within this model, the bot functions as a secondary code that allows for new meaning to arise in the process of communication and, consequently, for a bereaved person to renegotiate their new identity in the face of loss. 

Climate change represents an existential threat that places young people at an increased risk of mental distress, as their entire future is endangered by the rising global temperature. However, climate anxiety, mental distress about climate change, can also coexist with positive emotions such as climate hope. This qualitative study explores the possibilities for social robots to foster climate hope and act as climate communicators based on their previously demonstrated potential to act as an engaging medium for important information. We wrote three scenarios in which a robot is described as talking with youth, each scenario emphasizing one of the central facets of climate communication: empathy, positive information and personal action. Emotional reactions to and opinions about the scenarios were evaluated with 42 groups of ninth graders (n=115, 42 group responses) through an online questionnaire. The results show that all scenarios elicited positive but also negative reactions. The reactions were the most mixed regarding the empathy scenario. Our findings suggest that a robot as a climate communicator could attract young people's interest, and that the robot should be designed to communicate by providing objective information and decision support for personal action rather than empathy. Engaging interaction with such a robot implies a certain degree of artificial intelligence in communication, even though the factual information the robot provides should be drawn from reliable and objective sources rather than be generated with AI. However, the sustainability implications of the concept require careful consideration.