Paper ID
3480
Description
We study how artificial intelligence (AI) can influence the drug development process in the global pharmaceutical industry. Despite considerable effort made in developing drugs, pharmaceutical firms experience declines in novelty for drugs they produced. As AI becomes an important general purpose technology (GPT), it could be used to address some known challenges in the drug development process. Using two large-scale datasets that contain detailed historical records of global drug development and patents, we identify AI-related patents to approximate firms’ AI capabilities and construct a relatively new similarity-based metric to measure drug novelty based on their chemical structure. We find that AI can primarily affect the earliest stage in drug discovery when tasks are heavily dependent on automatic data processing and reasoning. However, it may not necessarily help with the more expensive and risky clinical trial stages that require substantial human engagements and interventions. Additionally, AI can facilitate the development for drugs at the medium level of chemical novelty more than at the extreme ends of the spectrum. Our study sheds light on the understanding of the roles and limitations modern technology can have on drug development, one of the most complex innovation processes in the world.
Recommended Citation
Lou, Bowen and Wu, Lynn, "Artificial Intelligence and Drug Innovation" (2019). ICIS 2019 Proceedings. 34.
https://aisel.aisnet.org/icis2019/economics_is/economics_is/34
Artificial Intelligence and Drug Innovation
We study how artificial intelligence (AI) can influence the drug development process in the global pharmaceutical industry. Despite considerable effort made in developing drugs, pharmaceutical firms experience declines in novelty for drugs they produced. As AI becomes an important general purpose technology (GPT), it could be used to address some known challenges in the drug development process. Using two large-scale datasets that contain detailed historical records of global drug development and patents, we identify AI-related patents to approximate firms’ AI capabilities and construct a relatively new similarity-based metric to measure drug novelty based on their chemical structure. We find that AI can primarily affect the earliest stage in drug discovery when tasks are heavily dependent on automatic data processing and reasoning. However, it may not necessarily help with the more expensive and risky clinical trial stages that require substantial human engagements and interventions. Additionally, AI can facilitate the development for drugs at the medium level of chemical novelty more than at the extreme ends of the spectrum. Our study sheds light on the understanding of the roles and limitations modern technology can have on drug development, one of the most complex innovation processes in the world.