Reining In Unbridled AI Enthusiasm: Protecting the Integrity of Rehabilitation Science & Clinical Care.

Abstract

Enthusiasm for artificial intelligence (AI) applications in rehabilitation medicine has accelerated rapidly, from 22 publications in 2015 to 1,449 in 2025. This expansion reflects growing optimism that AI may enhance clinical efficiency, streamline research workflows, and support patient-centered care. Uncritical adoption carries significant risks for rehabilitation practice, research integrity, clinical workforce sustainability, health equity, and environmental impact. This commentary cautions against a hasty adoption of AI, and argues that responsible deployment in rehabilitation requires deliberate, ethical integration grounded in shared responsibility across clinicians, developers, healthcare systems, and researchers. For rehabilitation providers, ethical AI use demands grounding decisions in discipline-specific principles, understanding model capabilities and limitations, implementing transparent informed consent processes, ensuring data security, and reviewing AI outputs before clinical application. Clinicians must remain accountable for AI-generated recommendations while collaborating with developers to mitigate bias and inequities. Patient-facing applications require particular safeguards, including patient education and meaningful human oversight through regular auditing by trained clinicians. Developers should establish minimum standards for representative data curation, ongoing model updating, and quality assurance, while maintaining cultural responsiveness and transparency about uncertainty in recommendations. Healthcare systems must engage multi-stakeholder groups-including patients, clinicians, caregivers, and community representatives-to evaluate ethical, labor, societal, and environmental implications, prioritizing patient wellbeing over efficiency gains. Researchers must critically examine epistemological and ethical implications of AI methodologies while preserving theory-driven inquiry and community-based participatory approaches. These coordinated efforts are essential to ensure AI enhances rehabilitation while preserving core ethical principles, protecting patient rights, and promoting evidence-based equitable care.