Kourosh Davarpanah , 2025-05-14 14:54:00
The CEOs of Anthropic, DeepMind, and OpenAI regularly single out drug discovery as the area where AI will have the greatest positive impact over the next decade. The first generation of AI drug discovery startups have yet to deliver, but to quote Dario Amodei, CEO of Anthropic, the belief is that the next generation of “powerful AI could at least 10x the rate of these discoveries, giving us the next 50-100 years of biological progress in 5-10 years.”
If this Cambrian explosion really happens, we are not ready for it. Identifying orders of magnitude more drug candidates is only part of the battle. Silicon Valley tends to focus on discovery and brush off development, but until we have viable alternatives, clinical trials remain the only path to bring these discoveries to patients.
Clinical trials are the most expensive and time-consuming part of drug development, costing hundreds of millions of dollars and taking up to a decade. Unless we find a way to accelerate clinical trials, we will end up with a hundred times more drug candidates but no more drugs reaching the patients that need them.
The good news is AI isn’t just a tool for discovery. It can also transform development, but this will require a fundamentally different type of intelligence. Unlike discovery, you can’t just throw raw horsepower at trials and 10x your output. The ecosystem is too fragmented across too many stakeholders. You need AI agents that support patients, healthcare professionals, and trial sponsors through a process that is both complex and high stakes.
Sponsors, sites, and patients all have their own challenges and goals. Sponsors want to bring new drugs to market faster. Sites want to offer more trials to their patients. Patients want to make the best decisions for their health. Generative AI can now solve big, discrete problems for each stakeholder that were out of reach just a year ago. Sponsors can use AI to design better protocols. Sites can use it to identify eligible patients across all their active trials. Patients themselves are increasingly using chatbots to support their decision making and take control of their health.
But it’s not enough for AI to improve individual experiences in a silo. This approach consistently breaks down in practice. Sponsors use AI to generate a dream list of sites informed by decades of historical data, only to find out some sites aren’t even interested in the trial. Sites apply for a trial that precisely fits their patient needs, only to find out three months later that they weren’t selected by the sponsor. A patient expresses interest in a trial, then waits three weeks to successfully share their medical record with the site and find out if they’re even eligible to participate.
It’s a six-legged race. One stakeholder can’t sprint ahead and expect to be successful. Sponsor, site, and patient experiences are too interconnected; they need to operate in sync to drive real medical progress. The real change with AI will happen not when everyone adopts the technology, which is happening quickly, but when these AIs can actually communicate and coordinate with one another.
In the near future, we expect all stakeholders to have dedicated AI agents that can collaborate with each other without requiring standardized APIs or integrations. Here’s what this would look like. Research sites map each of their patient’s needs against all available trials. They automatically offer the right trial to the right patient at the right time. They onboard them within days, with personalized support to keep them engaged throughout the trial.
Imagine this at scale across tens of thousands of sites globally. Sponsors instantly see where each drug meets real patient needs. They prioritize programs with the highest unmet need and partner globally with sites that have eligible, interested patients. Trials fill in weeks, not years.
Development has lived in the shadow of discovery for a long time, but the reality is that discovery cannot succeed alone. It’s time to build AI agents that can collaborate across the entire research ecosystem to rapidly translate drug breakthroughs into real patient impact.
Photo: Yuuji, Getty Images
Kourosh Davarpanah is the co-founder and CEO of Inato. Under his leadership, Inato is building a platform that combines early planning, site selection, and enrollment to make clinical trials more efficient and inclusive.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
