AI-driven patent analytics systems employ natural language processing (NLP) and machine learning models to analyze patents. These systems can categorize patents based on specific criteria, such as technological domain, application area, and geographical region. They can also identify relationships between different patents and predict future trends.