The ongoing digital transformation has aroused hopes of climate protection potentials of information and communication technologies (ICT) within manufacturing industries. Environmentally beneficial effects may range from the dematerialization of products to energy and resource efficiency gains. However, at the same time, ICT consume energy and resources themselves and detrimental direct and indirect effects of digital technologies are coming increasingly into focus such as emissions related to their production and use. Accordingly, whether trade-offs or synergies between the use of digital technologies and environmental benefits exist is ambiguous ex ante and effects may be highly heterogeneous depending on the context. Thus, our analysis sheds light on the most important moderators of the relationship between ICT and energy use in manufacturing by conducting an exploratory analysis. We apply flexible tree-based machine learning algorithms on an official administrative panel data set collected between 2009 and 2017 on more than 25,000 manufacturing firms to measure effect heterogeneity. Very preliminary results suggests that the relationship is weakly heterogeneous and effect differences can be related to fossil fuel use and the energy price which is exogenous to the firm.