The Radio Neutrino Observatory in Greenland (RNO-G) aims to detect EeV energy neutrinos. Operating primarily in the 200-700 MHz range, RNO-G is sensitive to broadband signals originating from many sources, including thermal noise, galactic noise, and various types of anthropogenic and environmental backgrounds. Each of these backgrounds will have a unique event signature that can be identified through reconstruction, polarization, and frequency content. During the 2021 and 2022 deployment seasons, 7 of the envisioned 35 stations have been installed and are taking data. This data lends itself well for initial analyses and the characterisation of the classes of noise recorded by RNO-G. In this contribution, I will discuss the current status of background identification with RNO-G. In particular, I will present our efforts to detect noise from our Galaxy, which might be used as a standard candle for future calibration. I will also discuss our efforts to develop new types of event classification strategies, including anomaly detection and other machine learning tools.