AI Revolutionizes Pharmaceutical Production, But Not the Labs

AI and Drug Discovery: A Mixed Success

In the pharmaceutical sector, artificial intelligence (AI) has yet to revolutionize drug discovery, despite billions in investments. Diogo Rau, Chief Information and Digital Officer at Eli Lilly, emphasized that AI has not significantly improved the success rates of clinical trials. This statement comes as Eli Lilly invests heavily in partnerships with Nvidia and develops one of the most powerful supercomputers in the industry. Companies such as Roche, GSK, AstraZeneca, and Merck have also signed multi-billion dollar contracts with AI specialists in recent months.

However, analyst Trung Huynh from RBC remains skeptical about the improvement of clinical trial success rates through AI. Recursion Pharmaceuticals, a pioneer in the field, has yet to commercialize an AI-developed drug after 13 years of operation and even had to reduce its workforce by 20% last year. The company aimed to overcome the notorious 90% failure rate in drug development.

The Real Benefits of AI in the Pharmaceutical Industry

According to the Wall Street Journal, the true successes of AI in the pharmaceutical industry lie in optimizing administrative tasks and manufacturing. For instance, Eli Lilly has used machine learning to create a digital twin of its manufacturing process for tirzepatide, an active ingredient in Mounjaro and Zepbound. This has led to reduced production time and increased output.

There are a few early successes in drug discovery itself. Recursion designed an experimental cancer drug in just 18 months, compared to an industry average of four years. However, human trials remain lengthy. Despite these challenges, RBC estimates that AI could enable the U.S. pharmaceutical industry to save around $90 billion over the next five years. These savings primarily stem from improved manufacturing processes and administrative management, rather than major breakthroughs in research and development.

AI has enabled the creation of sophisticated digital models that optimize production chains, thereby reducing costs and the time required to bring products to market. These technological advancements allow pharmaceutical companies to better manage their resources and increase operational efficiency. However, despite these advancements, the path to a complete revolution in drug discovery through AI remains fraught with challenges, and initial expectations must be reassessed in light of current results.