Artificial intelligence (AI) systems have increasingly been adopted in various fields, including education, healthcare, law, and journalism, due to their ability to save time, reduce costs, and ease human efforts. The growing relevance of AI systems brings a need to prepare undergraduate students from diverse backgrounds to understand and use AI technologies productively and responsibly in their professional careers. However, in order to effectively introduce AI to non-computer science undergraduates, it is essential to investigate the necessary AI competencies these students need to acquire. Therefore, in this work, we have conducted semi-structured interviews with multidisciplinary higher education lecturers with AI knowledge to analyze which AI competencies are relevant to be included in undergraduate curricula for non-computer science students according to their perspectives. This article presents the findings of these interviews as well as the emergent list of competencies. The list covers various aspects of AI, ranging from introductory topics, such as the capabilities of AI systems, to more advanced theoretical knowledge and practical skills in data management and machine learning. Moreover, the list contains competencies related to responsible AI, including the ethical and social implications of AI. Our list extends prior work on AI competencies for non-computer science undergraduates.