Humaniora editorial
Open AccessOA2025
On the Ethical and Moral Dimensions of Using Artificial Intelligence for Evidence Synthesis
This article explores the ethical and moral challenges of using AI for evidence synthesis, questioning assumptions of efficiency and highlighting issues of moral agency, bias, and epistemic uncertainty. It argues for robust regulatory frameworks and institutional oversight to ensure ethical AI use in healthcare.
Soumyadeep Bhaumik· PLOS Global Public Health· 2025· DOI 10.1371/journal.pgph.0004348
The core problem
Artificial intelligence (AI) is increasingly being used in medicine and healthcare, including for evidence synthesis. However, there is a lack of articles specifically examining the ethical and moral dimensions of AI use for evidence synthesis. This article aims to fill this gap by employing the Adda (আড্ডা) approach from Bengali philosophy and culture—a form of oral exchange involving deep but conversational style discussion. Adda developed as a form of intellectual resistance against the cultural hegemony of British Imperialism and entails asking provocative questions to encourage critical discourse. The author questions the ethical assumption that AI enhances efficiency in evidence synthesis without empirical evidence, and examines challenges such as lack of moral agency of AI, bias amplification, epistemic uncertainty, impact on knowledge systems, and the need for oversight beyond technical evaluation.
Innovation
The article adopts the Adda approach, a traditional Bengali form of intellectual discourse that involves conversational yet deep discussion. This method is used to critically examine the ethical and moral dimensions of AI in evidence synthesis. The author draws on philosophical and cultural insights to structure the argument, moving beyond technical evaluations to consider broader ethical implications. The approach encourages provocative questioning and critical discourse, aligning with the spirit of Adda as a form of resistance and intellectual engagement.
Introduction
Artificial intelligence (AI) is increasingly being used in medicine and healthcare, including for evidence synthesis. However, there is a lack of articles specifically examining the ethical and moral dimensions of AI use for evidence synthesis. This article aims to fill this gap by employing the Adda (আড্ডা) approach from Bengali philosophy and culture—a form of oral exchange involving deep but conversational style discussion. Adda developed as a form of intellectual resistance against the cultural hegemony of British Imperialism and entails asking provocative questions to encourage critical discourse. The author questions the ethical assumption that AI enhances efficiency in evidence synthesis without empirical evidence, and examines challenges such as lack of moral agency of AI, bias amplification, epistemic uncertainty, impact on knowledge systems, and the need for oversight beyond technical evaluation.
Why it matters
The author discusses the ethical and moral responsibilities of governments, multi-laterals, research institutions, and funders in regulating and overseeing the development, validation, and conduct of evidence synthesis using AI. It is argued that industry self-regulation for responsible AI use is unlikely to address ethical and moral concerns. Instead, there is a need to develop legal frameworks, ethics codes, and bring such work within the ambit of institutional ethics committees. This would enable appreciation of the complexities around AI use for evidence synthesis, mitigate against moral hazards, and ensure that evidence synthesis leads to improvement of health of individuals, nations, and societies. The Adda approach facilitates this critical discussion, emphasizing the importance of ethical considerations in the deployment of AI in healthcare research.
Who should read this
Pembaca Research Digest dan praktisi rumpun ini.
Opening member content…