🚀 Science
AI halves the time to find new drugs, but clinical trials still take years
In 60 seconds: Drug companies such as Novo Nordisk, Sanofi and Dr Reddy's are using AI to find drug candidates, design molecules and run trials. Funding for AI drug discovery reached $8.4 billion in 2025, up from $4.1 billion in 2023. Early research that took four to five years can now take about a year, but clinical trials from Phase I to Phase III still take seven to eight years.
Why it matters
Faster discovery could bring new medicines sooner and cheaper. India, strong in making drugs for others, has a chance to move into original research.
Explain it simply
AI is like a smart assistant who tastes a thousand tea blends in a day and picks the ten best. You still have to serve those ten to customers for months before you know which one sells.
The 3-minute read
What happened
Companies are building their own AI tools and partnering with tech firms. Recent deals include AstraZeneca with Insitro (October 2026), Eli Lilly with Insilico Medicine (2026) and Novartis with Isomorphic Labs (2024).
Background
A new drug typically takes 10 to 12 years from early research to approval. So far, AI's gains are mostly in the early stages: finding targets, designing molecules and picking promising candidates.
How it works
Sanofi says its AI tools sped up clinical trial site selection by 60% and cut mRNA design time by half. Dr Reddy's uses an in-house platform, Aurigene.ai, in its contract research business.
What's next
The global AI drug discovery market is estimated to grow from $2.3 billion in 2025 to $13.8 billion by 2033. The big test is whether AI can also shorten clinical trials.
What it means for you
For students, the overlap of biology, chemistry and computing is opening new careers in research, data science and pharma.
Key numbers
- $8.4 bn: AI drug discovery funding in 2025
- ~1 year: Early discovery time with AI (was 4–5)
- 7–8 years: Phase I to III trials, unchanged
- $13.8 bn: Projected market by 2033
Both sides
- Drug makers: AI helps prioritise experiments and use resources more efficiently.
- Indian start-ups: Trials remain the bottleneck, so small firms must partner with big companies to share the cost.
Key terms explained
- Clinical trials
- Tests of a new medicine on volunteers, in phases, before it can be approved.
- Drug candidates
- Molecules that show promise and are chosen for further testing.
The story so far
- 2020: AI helps spot an existing drug as a Covid-19 treatment. In 2020 UK-based BenevolentAI identified baricitinib, an existing drug developed by Incyte and licensed to Eli Lilly, as a potential Covid-19 treatment, showing how AI can find new uses for old medicines.
- 2024: Nobel Prize in Chemistry for protein design and structure prediction. The 2024 Nobel Prize in Chemistry went to David Baker for computational protein design, and to Demis Hassabis and John Jumper of Google DeepMind for AlphaFold, which predicts protein structures.
- Today: AI speeds drug discovery; trials still slow
Exam notes (UPSC, SSC, Banking, PSC)
- Clinical trials test a new drug in people in phases: Phase I (safety), Phase II (effectiveness and dose), Phase III (large-scale confirmation).
- In India, new drugs and clinical trials are regulated by the CDSCO under the New Drugs and Clinical Trials Rules, 2019.
- AlphaFold, an AI system by Google DeepMind, predicts the 3D structure of proteins.
- mRNA is the molecule that tells cells which proteins to make.
Likely exam question: How is artificial intelligence changing drug discovery, and what are its limits?