The pharmaceutical industry is reassessing its approach to artificial intelligence in drug discovery as leaders acknowledge significant gaps between early promises and clinical outcomes. Mark DePristo, CEO of BigHat Biosciences, cautions that widespread misconceptions about AI capabilities risk undermining legitimate progress in computational drug development.
Despite annual global investments reaching $50 billion, AI-designed drugs demonstrate only a 3% success rate in Phase II and later clinical trials. Current machine learning systems excel at pattern recognition and specific applications such as target identification and compound optimization, yet struggle with the fundamental complexity of biological systems.
Industry executives are now emphasizing a more measured perspective: AI serves as a valuable tool within a larger pharmaceutical ecosystem that remains dependent on traditional research methods, regulatory expertise, and manufacturing capabilities. This collaborative approach reflects a maturation in the field toward realistic, evidence-based applications.
Read the full article on GMJ Newsroom.
Was this article helpful?
