Artificial intelligence and machine learning in pharmaceutical research and healthcare: Ethical challenges and a...

Artificial intelligence (AI) and machine learning (ML) are transforming pharmaceutical research and healthcare by enabling analysis of large-scale biomedical data and supporting data-driven decision-making. However, their rapid integration has introduced significant ethical, governance, and implementation challenges that remain insufficiently synthesized within a unified framework.

This work aims to synthesize the central ethical challenges and paradoxes associated with AI and ML in pharmaceutical and healthcare contexts and to develop a structured, actionable governance framework to guide responsible implementation.

We conducted a narrative review using structured searches of PubMed/MEDLINE and Google Scholar (January 2015-February 2026). We screened peer-reviewed, English-language publications addressing AI/ML applications or associated ethical considerations and synthesized evidence thematically.

A total of 127 studies were synthesized, revealing interconnected ethical challenges, including bias, accountability, privacy, and equity, affecting patient safety and trust. The literature clustered across pharmaceutical research, clinical decision support, and governance, informing a structured, ethics centered framework with three pillars including technical excellence and safety, robust governance and trust, and human-centered values, along with cross-cutting enablers.

Successful integration of AI/ML into pharmaceutical research and healthcare requires addressing its inherent ethical paradox. By linking interconnected challenges to a structured, actionable governance framework, this study provides practical guidance for implementing trustworthy, equitable, and human-centered AI, with direct implications for patient safety, clinical practice, and policy.