The design and implementation of AI solutions or software solutions that utilise AI requires a combination of technical, ethical, and creative skills. Technical implementation skills are essential for creating valuable, high-quality, and user-friendly AI systems. A comprehensive understanding of the capabilities and limitations of AI is also crucial, as is the ability to reflect on ethical implications and make informed decisions about the suitability of AI for a given application. Additionally, a good understanding of future users and a creative approach are vital for the development of AI systems that are both effective and acceptable to users. It is uncommon for a single developer to possess all of these skills, which underscores the necessity for interdisciplinary collaboration between computer scientists, ethicists, and designers. Such collaboration must be taught and practised, and a framework must be established to ensure its success. The objective of this paper is to present an approach to the joint training of computer scientists and designers that builds on a methodological framework combining elements of Design Thinking with those of EduScrum. This approach incorporates the results of interdisciplinary collaboration and coaching. The approach was tested in a project in which students of computer science and visual communication collaborated to design and implement an exhibition that would facilitate critical reflection on AI and the responsibility of developers. The project was evaluated using a mixed-method approach, comprising qualitative interviews, evaluation of learning reflection and recorded videos. This paper presents the results pertaining to the methodological framework.