<p>Feedback on clinical performance is necessary to enhance the skill of therapists in training, yet it is very difficult to obtain due to the burden of time involved and scarcity of skilled resources to provide it. The current study examined the use of a commercially available artificial intelligence (AI) tool as a means of providing routine, timely, focused feedback in a doctoral course on clinical intervention. Successive cohorts of the class, one of which received AI feedback each week and one of which did not, were compared in terms of performance trajectory over the course of a semester. In both cases, students submitted recordings of weekly role-play homework assignments, which were immediately coded by AI software to produce standardized clinical feedback in the second cohort. Multilevel modeling indicated that the group that received weekly AI feedback exhibited significantly lower capability at baseline but improved at a significantly higher rate such that their skill was rated significantly higher at the end of a semester. The percentage of therapist speaking time did not differ across groups at baseline and did not result in substantial changes to the overall results when included as a time-varying predictor in the analytic model. Software-based feedback may have implications for graduate clinical psychological training, which could be explored further with future randomized trials. The applied usage of such software in clinical graduate training is discussed, with an emphasis on enabling more frequent and immediate feedback in comparison to the now-standard model that requires intensive human inputs.</p>