Sir, randomised controlled trials (RCTs), as the cornerstone of evidence-based dentistry, are considered the most reliable primary studies for evaluating the effectiveness of dental interventions. However, researchers are increasingly exploring new approaches to address the limitations of RCTs, such as their high cost, time-consuming nature, limited external validity, reliance on surrogate outcomes, and patient recruitment difficulties.1

Recent advances in artificial intelligence (AI), particularly deep learning models pre-trained on large patient datasets, have the potential to mimic RCTs for treatment effect estimation. Liu et al. developed CURE, an AI framework pre-trained on anonymised data from millions of patients, which successfully predicted the efficacy of stroke prevention treatments in individuals with heart disease, achieving results comparable to those of established RCTs.2

This technology holds exciting potential for dentistry. Imagine an AI model trained on massive datasets of dental patients, including demographics, lifestyle, medical history, treatment details, and long-term outcomes to identify patterns and predict the most effective treatments for individuals. This personalised approach could revolutionise dental care by optimising treatment selection based on each patient's unique profile. It could enhance risk assessment to identify people at higher risk for conditions like caries, periodontal disease or medication-related osteonecrosis of the jaw, enabling early preventive measures and reducing the need for invasive treatments.

Furthermore, AI algorithms can streamline clinical trials by accelerating patient enrolment and facilitating predictive modelling of oral health problems and outcome evaluation.1 Interestingly, AI can also generate synthetic patient data for virtual clinical trials, reducing costs and protecting the privacy of real patients.3 If synthetic data accurately reflects real populations, it could be used pre-clinically to test new treatments before human trials.

However, like CURE, AI effectiveness in dentistry relies on high-quality, complete training data. This highlights the importance of acknowledging the limitations of current AI technology in dentistry.1,2 Furthermore, the complex interplay between biological, behavioural, and social factors in oral health makes it difficult to assess the effectiveness of treatment using AI alone.

Although AI is not expected to completely replace RCTs, it complements them and offers a promising tool for personalised dentistry. Combining the analytical power of AI with the established rigour of RCTs can accelerate advances in dental care and ensure optimal patient outcomes.