A company trains an AI hiring tool on ten years of resume and hiring outcome data. The system learns to score new candidates based on patterns from those past decisions. Which type of learning is this?

An AI system clusters job applicants into groups based on patterns it identified in the data, without anyone specifying what categories to look for. A hiring manager cannot fully explain what the clusters represent. Which type of learning produced this outcome?

A recruitment AI is designed to recommend candidates that hiring managers are likely to accept. Over time it learns to surface fewer candidates with career gaps, because managers have historically accepted them less often. The system was never explicitly told to penalize career gaps. What best explains this behavior?