Pharmaceutical companies are pouring money into artificial intelligence for drug discovery, and the industry is now confronting a harder question: do the medicines actually work?
According to Genetic Engineering and Biotechnology News, pharma is "racing to scale AI" as billions of dollars flow into drug discovery. The push is global. Fortune India reports that Minister of State for Health Anupriya Patel said India's domestic pharmaceutical industry is deploying AI for drug discovery.
The technical frontier is shifting, too. A GlobeNewswire item published by the Manila Times argues that most AI drug discovery today reads the genetic code — the ordered letters of sequence data that modern computational biology is built on — while the next wave aims to learn what that code actually does. MIT Technology Review has covered how AI helps scientists design the next generation of medicines, and Phys.org published a Q&A on how researchers are using AI to speed up discovery and development. Nature has gone a level up, publishing research on the structural characteristics and evolutionary trajectories of knowledge recombination across the AI-driven drug discovery field itself.
But the enthusiasm comes with a caveat that several outlets are now foregrounding. The Economic Times' pharma vertical frames it bluntly: AI has cracked the drug discovery code, but clinical proof remains the missing link — AI tools are shortening timelines, yet efficacy and safety still have to be validated in human trials. The Pharma Letter calls this AI drug discovery's "year of evidence," and Pharma Voice is asking directly how AI-discovered drugs are faring in the clinic.
Why it matters: AI can generate drug candidates far faster than traditional methods, but a molecule only becomes a medicine after it proves safe and effective in patients — and that verdict, not the speed of discovery, will determine whether the billions being invested pay off.