Artificial intelligence in human-assisted reproduction: a paradigm shift still in search of clinical impact

Artificial intelligence holds greater promise in domains such as gamete assessment, automated data extraction, and personalized outcome prediction, where it may enhance treatment management, counselling, and decision-making for both clinicians and patients. Realizing this potential requires a strategic shift in research priorities and rigorous adherence to shared standards, including model transparency, uniformity, external validation, and benchmarking against established clinical workflows. Without such recalibration, artificial intelligence risks becoming a hyped technology with limited clinical relevance in assisted reproduction.