The thermo-mechanical load collective prevailing during machining significantly influences the wear behavior of coated carbide tools and thereby impacts the productivity and sustainability of many industrial value chains. AlCrVY(O)N coatings show a significant potential for reducing tool wear due to their lower friction and enhanced thermal stability at elevated temperatures compared to conventional coatings. However, due to the complex interplay between friction, temperature, and wear mechanisms, further development of such coating systems remains essential. Additionally, detecting tool wear using black box approaches based on process data provides valuable insights into the wear process and coating behavior during machining operations. In this study, AlCrVY(O)N coatings were benchmarked against TiAlN coatings under varying cutting parameters to evaluate their performance. The variation in cutting parameters allowed for an investigation of coatings under different load levels. Sensor data collected during machining was used to train a long short-term memory autoencoder (LSTM) for analyzing the tool wear. Fundamental investigations revealed that the applied coatings exhibit a homogeneous chemical composition and high hardness values around 35 GPa. However, cutting tests showed that the temperatures on the rake face were not significantly reduced when using AlCrVY(O)N coatings compared to TiAlN coatings. The developed LSTM-autoencoder successfully identified tool wear by analyzing differences in the reconstruction error and latent space representations caused by variations in coating properties.