PI3K pathway is heavily emphasized in cancer where PIK3CA, which encodes for the p110 alpha subunit of PI3K alpha, presents itself as the second most common mutated gene. A lot of effort has been put in developing PI3K inhibitors, opening promising avenues for the treatment of cancer. Among these, PI3K alpha specific inhibitor alpelisib was approved by FDA for breast cancer and other alpha-isoform specific inhibitors such as inavolisib and serabelisib reached clinical trials. However, the mode of action of these inhibitors on mutated PI3K alpha and how they interact with mutant structures has not been fully elucidated yet. In this study, we are revealing the calculated interactions and binding affinities of these inhibitors within the context of PIK3CA hotspot mutations (E542K, E545K and H1047R) by employing molecular dynamics (MD) simulations. We performed principal component analysis to understand the motions of the protein complex during our simulations and also checked the correlated motions of all amino acids. Binding affinity calculations with MM-PBSA confirmed the consistent binding of alpelisib across mutations and revealed relatively higher affinities for inavolisib towards wild-type and H1047R mutant structures in comparison to other inhibitors. On the other hand, E542K mutation significantly impaired the interaction of inavolisib and serabelisib with PI3K alpha. We also investigated the structural relationship of the natural ligand ATP with PI3K alpha, and interestingly realized a significant reduction in binding affinity for the mutants, with potentially unexpected implications on the mechanisms that render these mutations oncogenic. Moreover, correlated motions of all residues were generally higher for ATP except the H1047R mutation which exhibited a distinguishable reduction. The results presented here could be guiding for pre-clinical and clinical studies of personalized medicine where individual mutations are a strong consideration point.