Drugmakers Use AI Everywhere, but Few Think It Will Actually Work
Faster drug research doesn't mean better medicines — here's the catch.
If you think AI is just for chatbots and self-driving cars, think again. A new survey from Citi Research found that almost every major US pharmaceutical company has adopted AI in its research labs. In fact, 72% of tech-strategy executives said their companies were either scaling up or had fully deployed AI across drug development. Not a single company said it had no plans to use the technology. The promise? AI could speed up the hunt for new medicines and save up to $26 billion globally in discovery costs.
But here's the twist. Despite all that enthusiasm, only 24% of those same executives believe AI will lead to significantly better drugs that are more likely to win approval. More than half expect just a 'moderate' improvement, not a revolution. As one Citi analyst put it, the biggest risk isn't that AI fails to accelerate discovery — it's that faster discovery doesn't translate into drugs that actually work better for patients.
Why the skepticism? The drug industry's toughest problems aren't just about finding molecules faster. A third of executives said human biology, clinical judgment, and execution will remain major bottlenecks no matter how powerful the AI gets. After all, a shorter race to the finish line doesn't help if the race itself is heading toward the wrong target. Even an AI-discovered drug still has to survive years of human trials, where most candidates fail for reasons that have little to do with speed.
For you, this means AI's real benefit may be lower research costs and shorter timelines for some treatments — but don't expect a sudden flood of miracle cures. The gap between tech hype and real-world outcomes is still wide, and patients may have to wait longer than the headlines suggest.
- 72% of surveyed US pharma companies are already scaling or have fully scaled AI in research — with zero firms saying they'd skip it.
- Only 24% of executives expect AI to significantly boost drug success rates; 56% see only a moderate gain and 34% say human biology and clinical judgment will still bottleneck progress.
- AI may cut up to $26 billion in drug-discovery costs, but faster research won't automatically mean safer or more effective medicines for patients.
Why It Matters
AI could change how your future medicines are discovered—but faster research doesn't equal more cures, so know the real score.