In this innovative, first-of-its-kind study, over 7,000 daily sunspot drawings spanning a longstanding 52-year period from 1971 to 2023 were analyzed using artificial intelligence to classify active sunspot regions. A convolutional neural network was developed and trained on this rare dataset of hand-drawn solar observations from the Astronomical Institute of the Czech Academy of Sciences. The model successfully categorizes sunspot groups according to the McIntosh classification system. This research shows that machine learning is a valuable tool in solar astronomy, enabling more accurate and automated classification of solar activity. It opens the door to improved space weather forecasts, whether detecting dangerous eruptions or beautiful auroras.