The paper presents a series of research focusing on the detection of welding imperfections using Deep Learning algorithms and artificial intelligence (AI) technology. This research direction falls within a highly interesting field, both for companies and research institutes or universities. According to the introductory study presented in the paper, a significant increase in the impact of AI on industrial activities is estimated, particularly on non-destructive testing and examination (NDE) methods, as well as for an increasingly wide range of material processing technology, including in the field of welding materials.