Abstract
Artificial intelligence (AI) holds great potential for minimally invasive surgery, with fields of application ranging from interdisciplinary treatment stratification through preoperative planning up to active decision support in the operating room, which are the focus of this article. Artificial neural networks for analysis of surgical video recordings could enhance surgical safety, efficiency and planning. High-quality, diverse (meta)data are essential for such AI applications but the annotation, training and validation present complex demands. Despite technological advances, the clinical implementation often fails due to a lack of data standardization, insufficient infrastructure, regulatory barriers and ethical uncertainties. Many models remain black boxes, which hinders acceptance and trust among medical professionals. In addition, AI systems need to be robust, transparent and practically integrable into clinical workflows. Stringent data collection strategies, privacy-preserving learning methods, explainable AI and human-in-the-loop approaches are critical to facilitate clinical translation. Regulatory framework conditions, such as the General Data Protection Regulation, the EU Medical Device Regulation and the EU AI Act, will require further legal refinements to address the specific needs of medical AI applications and interventions, to facilitate the safe adoption of interdisciplinary assistive technologies in the operating room that meaningfully support surgical practice.