With the publication of the Fintech Research Paper Dimensionality reduction techniques to support insider trading detection, the study initiated by CONSOB in collaboration with the Scuola Normale Superiore di Pisa on the solutions that new technologies, AI-applications based, can offer to support market supervision is expanding. Specifically, the study - based on an anonymized dataset - addresses the problem of identifying potential cases of insider trading and proposes a different methodological approach than previous studies that have made use of unsupervised machine learning techniques: in this case, in fact, the technique of decomposition and subsequent reconstruction of a time series of data through "principal components" (PCA, Principal Component Analysis) is applied, in relation to the positions taken by groups of investors in a given stock in the vicinity of a price-sensitive event.