Machine learning (ML) has seen increased use for Earth observations in recent years. Examples of such applications include estimating crop yield, mapping flood areas, and classifying settlements and buildings. Over the last decade, there has been tremendous progress in developing ML methodologies for a variety of Earth Science applications, however, there is progress to be made in applying ML for both environmental and development challenges. This session calls for research that applies ML, in data-scarce regions, to address environmental development challenges as well as data democracy, model access, cloud computing, and capacity-building efforts. This session aims to bring together scientists, non-profits, changemakers, and thought leaders in an effort to build collaboration, grow capacity to utilize cutting edge technology and work towards using ML for creating a positive impact for the planet in developing regions.