Despite numerous studies examining the impact of moral vs. nonmoral political content on social media engagement, the specifics of how distinct moral messaging captivates public attention remain unexplored. Scrutinizing over 10,000 original tweets from 2018 US Senate election candidates, the present work addresses this gap through natural language processing combined with machine and deep learning to probe (a) Are certain types of moral messaging among US politicians more effective in garnering attention?; and (b) If so, does this pattern differ among Democrats and Republicans online? While an unequal distribution emerged, wherein morally charged content elicited more responses than nonmoral rhetoric on both sides of the political aisle, the results indicate varying sensitivities between the factions to certain moral messaging, with Democrats being driven not only by care/harm and fairness/cheating but also by those signaling betrayal, subversion, or even degradation, whereas Republicans were captivated by a broader spectrum of concerns, including care/harm, fairness/cheating, loyalty/betrayal, authority/subversion, and purity. Given that social media is becoming increasingly pivotal in politics today, such findings advance our understanding by untangling how advanced methodologies can dissect the intricacies of moralities in the digital realm, shedding light on the potential dynamics and strategies through which Democratic vs. Republican candidates adopt to expand their reach via social media politicking.