A striking evidence gap has emerged in pediatric artificial intelligence: currently, zero randomized controlled trials exist to evaluate AI tools specifically for children’s developmental outcomes, despite these technologies being widely deployed in educational and healthcare settings.
This statistic underscores a fundamental problem in how AI applications are brought to market for pediatric use. While developers invest heavily in engagement metrics—with current industry focus allocating 95% of evaluation efforts toward user retention rather than developmental benefits—the scientific community has produced no rigorous clinical evidence demonstrating that these tools actually improve children’s cognitive, linguistic, or social development.
Dr. Dua Hassan’s analysis reveals this concerning disparity: the same children using AI tools daily in classrooms and homes are the subjects of zero adequately designed trials measuring real developmental outcomes. This represents a significant departure from how other pediatric interventions are regulated and deployed, raising important questions about clinical oversight and the application of evidence-based medicine principles to digital health technologies for children.
Read the full article on GMJ Newsroom.
Was this article helpful?
