This scoping review of randomized controlled trials (RCTs) on artificial intelligence (AI) in clinical practice highlights an increasing interest in AI across various clinical specialties and regions. The USA and China lead in the number of trials, with a focus on deep learning systems for medical imaging, particularly in gastroenterology and radiology. Most trials (81% of 86) report positive primary endpoints, mainly related to diagnostic yield or performance. However, concerns remain about the generalizability and practicality of these results due to the predominance of single-center trials, limited demographic reporting, and varying operational efficiency reports. Despite promising outcomes, publication bias and the need for more comprehensive research, including multicenter trials, diverse outcome measures, and improved reporting standards, are crucial. The review found that 79% of trials were conducted in a single country, with the USA and China accounting for the majority. Gastroenterology trials were notable for their uniformity, with all trials testing video-based deep learning algorithms in an assistive setup. Most trials evaluated deep learning systems for medical imaging, with a majority being video-based. AI systems operated on structured data, such as electronic health records, waveform data, and free text, using a mix of decision trees, neural networks, reinforcement learning, and other machine learning techniques. The review also found that 70% of trials reported favorable results for their primary endpoints, with a similar success rate in gastroenterology. However, some trials reported negative results, highlighting the need for more international collaboration and multicenter trials to ensure the generalizability of AI systems across different populations and healthcare systems. The review emphasizes the importance of patient-relevant outcomes and the need for comprehensive reporting and participant diversity to enhance the external validity of findings. The study also notes the potential of AI to improve care management, patient behavior and symptoms, and clinical decision-making efficiency. However, the true success of AI applications depends on their generalizability to target patient populations and settings. The review underscores the importance of addressing publication bias and the need for more research focusing on multicenter trials and diverse endpoint measures, especially patient-relevant outcomes. The findings highlight the growing interest in AI in clinical practice but also the challenges in ensuring its effectiveness and generalizability across different contexts.This scoping review of randomized controlled trials (RCTs) on artificial intelligence (AI) in clinical practice highlights an increasing interest in AI across various clinical specialties and regions. The USA and China lead in the number of trials, with a focus on deep learning systems for medical imaging, particularly in gastroenterology and radiology. Most trials (81% of 86) report positive primary endpoints, mainly related to diagnostic yield or performance. However, concerns remain about the generalizability and practicality of these results due to the predominance of single-center trials, limited demographic reporting, and varying operational efficiency reports. Despite promising outcomes, publication bias and the need for more comprehensive research, including multicenter trials, diverse outcome measures, and improved reporting standards, are crucial. The review found that 79% of trials were conducted in a single country, with the USA and China accounting for the majority. Gastroenterology trials were notable for their uniformity, with all trials testing video-based deep learning algorithms in an assistive setup. Most trials evaluated deep learning systems for medical imaging, with a majority being video-based. AI systems operated on structured data, such as electronic health records, waveform data, and free text, using a mix of decision trees, neural networks, reinforcement learning, and other machine learning techniques. The review also found that 70% of trials reported favorable results for their primary endpoints, with a similar success rate in gastroenterology. However, some trials reported negative results, highlighting the need for more international collaboration and multicenter trials to ensure the generalizability of AI systems across different populations and healthcare systems. The review emphasizes the importance of patient-relevant outcomes and the need for comprehensive reporting and participant diversity to enhance the external validity of findings. The study also notes the potential of AI to improve care management, patient behavior and symptoms, and clinical decision-making efficiency. However, the true success of AI applications depends on their generalizability to target patient populations and settings. The review underscores the importance of addressing publication bias and the need for more research focusing on multicenter trials and diverse endpoint measures, especially patient-relevant outcomes. The findings highlight the growing interest in AI in clinical practice but also the challenges in ensuring its effectiveness and generalizability across different contexts.