With the increased prevalence of multidrug-resistant Gram-negative bacteria, the use of colistin and other last-line antimicrobials is being revisited clinically. As a result, there has been an emergence of colistin-resistant bacterial species, including Acinetobacter baumannii and Klebsiella pneumoniae. The rapid identification of such pathogens is vitally important for the effective treatment of patients. We previously demonstrated that mass spectrometry of bacterial glycolipids has the capacity to identify and detect colistin resistance in a variety of bacterial species. In this study, we present a machine learning paradigm that is capable of identifying A. baumannii, K. pneumoniae and their colistin-resistant forms using a manually curated dataset of lipid mass spectra from 48 additional Gram-positive and -negative organisms. We demonstrate that these classifiers detect A. baumannii and K. pneumoniae in isolate and polymicrobial specimens, establishing a framework to translate glycolipid mass spectra into pathogen identifications.