AI in Drug Repurposing: Opportunities, Challenges, and Commercial Realities

AI in Drug Repurposing: Opportunities, Challenges, and Commercial Realities

Drug repurposing-finding new uses for approved or investigational drugs-has captured growing attention, especially as AI and machine learning revolutionize the process. By mining vast, multi-source datasets, AI can rapidly identify drugs with potential off-target effects or new clinical actions, moving repurposing from chance to a systematic, data-driven strategy.

Traditionally, such discoveries often occurred serendipitously- through off-target effects or unexpected biological actions. However, with the application of AI/ML methods, this process is becoming far more systematic and data-driven.

Many drugs exhibit polypharmacology, acting on multiple targets or pathways. AI can now mine vast, heterogeneous datasets — from chemical, genomic, and clinical sources — to uncover hidden drug–target–disease relationships. Models ranging from classical ML to deep learning are being used to power this revolution, and tools like DrugRep and RepurposedDrugs have emerged to support it.

The appeal is obvious:
✅ Reduced timelines: 3-12 years vs. 10–15 years
✅ Lower costs: <$500M vs. >~$2.5B
✅ Lower risk: 25-70% failure vs. 90–95% in de novo discovery

(https://www.technologynetworks.com/drug-discovery/articles/drug-repurposing-strategies-challenges-and-successes-384263)

However, the commercial reality is more complex.
Repurposed drugs often face weaker IP protection (e.g., method-of-use patents that may last only ~3 years). Moreover, if repurposing is pursued by a different company than the original developer — often after patent expiry — the return on investment is limited.

Since most late-stage clinical trials are costly, pharma companies may hesitate to fund them without strong IP incentives. That means the real progress in AI-driven drug repurposing may depend on government or philanthropic funding — where the focus is on affordable and faster access to medicines, not just profitability.

In short, AI/ML methods can accelerate drug repurposing scientifically — but the challenge lies in making it commercially sustainable.