Generative Artificial Intelligence in Healthcare: Opportunities, Challenges, and Future Directions

Abstract

Generative artificial intelligence (GenAI) systems, which generate new text, images or even molecular structures based on patterns learned by the system are quickly finding their way into the clinic and the lab. The current paper summarizes work (2023-2025) that was written recently with an India orientation to demonstrate areas where GenAI can realistically be beneficial and where it is crucial to be cautious. I review the applications in imaging, drug discovery, personalized care, and synthetic data using a literature-synthesis approach and discuss the primary ethical, legal, and workforce implications. The existing literature indicates that GenAI will be able to supplement the diagnostics, accelerate early-stage drug design, and be used to create privacy-preserving datasets, yet it brings threats of bias, explainability, and consent. In the case of India, which has vast untapped health demands and is gradually changing to an increase in digital health infrastructure, GenAI would particularly be beneficial should it be implemented within a robust governance structure and with human follow-ups. The last part outlines research agenda and policy actions that are likely to render the technology secure, fair, and useful in the Indian context.