Today, Artificial Intelligence (AI) is ubiquitous. There is no field, activity or job that does not yield to the implementation of AI-based systems for the, presumably, better performance of its functions. This can mean, on many occasions, that AI is applied when and where it is not necessary or that, if it is necessary, it is applied without taking into account requirements, conditions and terms that are absolutely essential to avoid unintended consequences. In particular, digitization and digital transformation, although often confused, are not only different terms but correspond to different states, processes and moments. It is practically impossible to undertake the digital transformation of any system, whether social, administrative, economic or physical, without first digitizing it. Despite this priority, the two tasks often overlap. The aim of this talk is to clarify the meaning of both concepts, pointing out the stages that each may entail and pointing out the unintended consequences that may arise from the use of AI in systems that have not been digitally transformed.