Ethics of AI in Patient Recruitment and Diversity

Abstract

By 2026, global clinical research has seen a shift from voluntary diversity programs to mandatory regulation. This report examines the ethical and technical issues at the interface of artificial intelligence (AI) and demographic representation in trial enrollment. In the U.S. under the Food and Drug Administration's Diversity Action Plans and in India's New Drugs and Clinical Trials (NDCT) Rules of 2026, sponsors must now demonstrate that trial groups reflect the real-world diversity of the intended patient population.^{1, 2, 3} AI tools like TrialGPT and DocTr have emerged as key solutions, reporting up to 98.7% reduction in screening time and a 58% improvement in matching quality through analysis of large electronic health record (EHR) datasets.^{4, 5, 6} However, these technologies introduce significant ethical concerns, including "black box" epistemic opacity and algorithmic bias that can institutionalize healthcare inequalities.^{7, 8, 17} A persistent digital divide also excludes groups that lack high-speed connectivity or digital literacy.^{9, 10} This paper argues that although AI is central to achieving large-scale diversity in clinical research results, it must be implemented within governance frameworks such as the EU AI Act and WHO ethical guidelines, which guarantee human oversight, transparency, and data justice.^{10, 11, 12, 13} We recommend adherence to PRISMA 2020 reporting standards to ensure methodological rigor in the evaluation of AI-based recruitment interventions.^{12, 14}

Keywords: Ethics, Clinical Trials, Patient Recruitment, Diversity, Artificial Intelligence, Bioethical