The ongoing ATLAS differential cross-section measurements of Drell-Yan scattering at high dilepton masses based on the LHC data collected in 2015-2018 will be presented. The sensitivity of these measurements to new phenomena in the SMEFT effective field theory framework will be discussed. To increase sensitivity to new phenomena, electron identification efficiency in ATLAS needs to be improved, as it is an important source of systematic uncertainty for these measurements. On-going work towards a Deep Neural Network based electron identification for future analyses will be shown.