<p>In this paper, we assess the aiming error of a stationary turret system given target location from automatic target recognition (ATR) systems assisted by artificial intelligence (AI). Research shows that AI based object detection algorithms significantly improve ATR performance in target detection and classification. However, there is no standard procedure for examining the accuracy in rotating the turret to an aimpoint provided by an AI enhanced ATR system. Therefore, we develop a methodology to assess the impacts of AI based object detection algorithms on the accuracy of a stationary controlled gun turret system. In our approach, three different object detection models provide target location of stationary targets as input to two Proportional-Integral-Derivative controllers that rotate the turret to the aimpoint in numerical simulations. We then perform three experiments with the target location data and simulation data that examine the correlations between a measure of AI error and standard object detection performance metrics, and the impacts on the accuracy of the controlled turret system. The results indicate that AI technologies could add significant error to turret control systems. In some cases, this error could be one to two orders of magnitude larger than the controller error depending on target range and controller design. Moreover, the results suggest that the confidence score may not correlate well with the object detection error, while the intersection-over-union, average precision and average recall correlate with turret accuracy with our system parameters. The results lay the groundwork for assessing the impacts of AI assisted ATR systems on turret accuracy and could be useful in the development of error budgets accounting for AI.</p>