Recursion’s incoming CEO Najat Khan
Recursion
After a challenging 12 years trying to use AI to create new medicines, cofounder Chris Gibson stepped down as CEO last week. Can R&D chief and new CEO Najat Khan turn it around?
In 2014, Recursion cofounder and CEO Chris Gibson promised to develop 100 drugs in 10 years, thanks to the company’s use of AI in drug discovery.
It’s been rocky. More than a decade on, Recursion still has not brought a single drug to market. Its shares have cratered 86% since its April 2021 IPO, giving it a market cap today of $2.2 billion. Earlier this year, Recursion culled half of its pipeline of drugs and in June it cut its workforce by 20%, the third round of layoffs since August 2024. And its revenue for the latest 12 months (through September 30) shrank by one-third to $44 million from $65 million while its losses ballooned by nearly 90% to a staggering $716 million.
But Najat Khan, who will take over as CEO on January 1 as Gibson steps back, likes doing hard things.
“A lot of people when I went to Recursion said, ‘Changing the rules about how medicine is made is really hard.’ I’m like, ‘All right, challenge accepted,’” Khan told Forbes in her first interview since accepting the top job. “Is it going to work? Is it not? It is so worth a try, and rising up to that challenge, because 20 years into my career, we still have a 90% failure rate.”
Khan joined Salt Lake City-based Recursion in April 2024 as head of R&D and chief commercial officer. Previously, she’d built Johnson & Johnson’s 250-person AI team from scratch as chief data science officer and global head of strategy for innovative medicine R&D. Khan has now been charged with translating the promise of AI into actual drugs after years of failed promises.
“Recursion was the first to the space, which was kind of incredible, but it’s been harder than people might’ve thought in the beginning,” Khan said. “It’s definitely hard. So how do you change that? How do you take this brilliant idea that’s been sitting out there for a long time and make it into something where we really start seeing drugs coming into fruition?”
Today, only 10% of drug candidates succeed. And even when they do, Khan said, it costs $3.5 billion and takes 10 years to get to market. Those numbers are part of the reason for ballooning healthcare costs in the U.S. as drugmakers seek to make back staggering R&D costs before their innovative treatments go off patent and can be made into generics. That’s key to understanding the industry’s hopes for AI drug development, which theoretically could bring medicines to the clinic faster at lower cost.
The big question for Khan is whether Recursion’s approach will be able to pull that off, a decade-plus into trying. From the firm’s New York City outpost, she pointed to Recursion’s internal pipeline as a promising indicator: currently half a dozen drugs in cancer and rare diseases. One AI-designed drug that it’s working on is for a hereditary genetic disorder in which people get hundreds of polyps in their gastrointestinal tracts that almost inevitably lead to colon cancer. Khan expects to have new clinical data on that therapy within weeks.
She also pointed to milestone payments it had received from pharmaceutical companies for partnership work. That includes a recent $30 million payment from Roche and Genentech, its subsidiary, for delivering a whole-genome map of a specialized type of immune cells designed to accelerate treatments for neurological diseases.
“I’m seeing the green shoots of success,” she said.
For the latest healthcare news, sign up for Forbes’ InnovationRx newsletter.
This map is part of Recursion’s attempt to decode biology by gathering and analyzing massive amounts of data about genes, proteins and real-world patients.
“You can’t create drugs until you have this map and this data,” she said. In biology, the data has been so rudimentary, fragmented and problematic that companies that want to use AI for drug discovery need to gather it together from scratch. “You have to start at the very, very beginning,” she said.
That means there’s still a long road to a real payoff in AI-designed drugs. In a recent research report, Jefferies analyst Dennis Ding called Recursion’s drug discovery platform “novel and promising,” but noted that validation of it may “take some time” with clinical readouts “difficult to interpret.” His rating on the stock: Hold.
MORE FROM FORBES
Source: https://www.forbes.com/sites/amyfeldman/2025/11/17/recursions-incoming-ceo-needs-to-prove-that-ai-drug-development-can-pay-off/
