Utilization of Artificial Intelligence in the Discovery of Indonesian Herbal Medicines: Opportunities and Challenges

Abstract

This study examines the potential use of artificial intelligence (AI) in the development of Indonesian herbal medicine. Indonesia, as a megabiodiversity country with more than 6,000 species of medicinal plants, has great potential in the development of herbal medicine, but is constrained by a long research process, high costs, and low success rates. The research method uses a descriptive qualitative approach through literature studies with content analysis techniques. The results of the study indicate that AI integration can accelerate the drug development process, reducing the time from 10-15 years to a few months, and reducing costs from USD 2.8 billion to USD 500 million - 1 billion. The ConvNeXt model achieved 92.8% accuracy in the classification of Indonesian medicinal plants, proving its effectiveness as a tool for documentation and preservation of biodiversity. The AI-based virtual screening technique successfully predicted the pharmacological potential of bioactive compounds without requiring early in vitro/in vivo tests. Challenges in implementing AI include limitations of integrated local databases, gaps in multidisciplinary collaboration, inadequate computing infrastructure, and ethical aspects related to indigenous peoples' rights and Intellectual Property Rights. This study recommends strategies to strengthen the AI ​​research ecosystem through the development of a national database, strengthening multidisciplinary collaboration, increasing infrastructure and human resource capacity, strengthening regulations, and public advocacy and education.