Robots are becoming more and more present in many places characterising daily life. In these contexts, they appear more social and similar to humans in their behaviours. Recent estimates indicate that a $6 billion market (or more) in people-sized-and-shaped robots is achievable in the next 10 to 15 years in order to address problems such as the projected manufacturing labour shortage and global elderly care demand. Social robots can communicate with high- level dialogues, i.e. the ability to ask questions and use dialogue to create engagement in the users, establish and maintain social relationships with users, use natural cues such as gaze, gestures, and movements, exhibit a distinctive personality and character, learn and develop social competencies. Thus, humanoid robots have great social potential because their human-like appearance and behaviour can stimulate and engage users, even those who are not used to interacting with technology. Indeed, such robots can attract people's attention, engage them, and improve the user experience (UX) by expressing emotions, communicating through high-level dialogues, using natural cues, and exhibiting social skills and distinctive personalities. Furthermore, these capacities can allow the robots to be employed to interact more naturally and socially rather than be considered mere instrumental tools. Robots can be an important addition not only for performing repetitive tasks but also for having a companion that can help in improving quality of life and emotional states. In anticipation of the broad impact that these emerging technologies will have on our lives, a reflective and systematic consideration is necessary that leverages their full potential in terms of user experience.In order to achieve such benefits artificial intelligence methods and techniques can play an important role. Recently, researchers have focused on robots that can adapt their behaviour to various human conditions and needs to facilitate natural interaction. In particular, Machine Learning (ML) techniques for adaptive social robots are finding more and more success. Adaptive robot interactions are important for providing comfortable and effective interactions with humans, fostering meaningful communication and building trust between users and robots. For example, Reinforcement Learning (RL) can enable robots to learn from their interaction with the environment and makes it possible to adapt and optimize robotic policies for different users. In addition, the advent of Large Language Models opens new opportunities for rapidly obtaining personalised interactions. In general, each individual may have different personalities, preferences, cognitive abilities, engagement, and attention levels in interacting with robots during task performance. For example, personalized and tailored robotic assistive systems can establish a productive interaction with the user, improving the effects of a therapy session. Adaptive robot interactions are necessary to provide a comfortable and effective interaction with users. In this perspective, we will also investigate the best way to integrate conversational abilities in the humanoid robot also exploiting the possibilities of Large Language Models, such as ChatGPT. Another opportunity for improving the personalisation of the interaction is the use of recommendations. During the conversation with the users, the recommendation system can acquire and understand their preferences, which will be used during later interactions to personalise the experience further.Based on such concepts the two groups led by F.Paternò at CNR and A.Krueger at DFKI aim to activate a collaboration to develop a research agenda for them and the relevant national and international communities and inform the work they carry out in the currently active projects.