This paper explores the potential of a modern AI-based photometric stereo model for detecting and assessing tool wear in machining. A retrofitted microscope is used to scan a test workpiece with a well-defined geometry, allowing for a quantitative evaluation of the resulting normal map. Additionally, a qualitative analysis is conducted on a scan of an end mill. Beyond normal maps, the study discusses the integration of additional Bidirectional Reflectance Distribution Function (BRDF) maps for the assessment of tool wear.