Chris Gibson built the pioneering AI biotech Recursion Pharmaceuticals on a foundation of bold promises. But after more than a decade at the helm, Gibson stepped down as CEO in November 2025. And his replacement, Najat Khan, Ph.D., is laser-focused on one thing: turning the amorphous potential of AI into tangible products for patients.
“The companies that are really going to be able to transition from promise to proof to product are going to be the ones that are the winners,” Khan told Fierce Biotech.
Today, in its quarterly earnings, Recursion shared what Khan considers another proof point for Recursion’s AI-powered drug development engine. The biotech’s cancer candidate REC-1245 has posted an early cut of safety and pharmacokinetic data, showing no dose-limiting toxicities and engagement with its target, the never-before-drugged RNA-binding protein 39 (RBM39).
“We’re really encouraged by the safety and tolerability profile that we see to date,” Khan said. Recursion is now “rapidly” enrolling patients in higher dose arms of the study, she said, which should lead to the first efficacy data later this year.
“We expect to have exposures consistent with tumor regression in mice in the next two to three dose levels,” Khan told Fierce.
REC-1245 is designed to degrade RBM39, which is overactive in many cancers and which Recursion thinks plays a role in genomic instability that can drive tumor progression.
“We made this degrader start to finish, all the way from the insight about the target and into the IND-enabling studies, in 18 months,” Khan said. REC-1245 joins Recursion’s lead program REC-4881, for the rare disease familial adenomatous polyposis, as “proof-of-concept” for Recursion’s platform, she added.
REC-4881 “was one of the earliest programs for the platform,” Khan explained. “We’re really building on layers of evidence. That gets me excited and energized about what we have ahead.”
Discussions with the FDA about a registrational trial for REC-4881 are underway and, according to Khan, are going well.
Related
But getting to this point took some growing pains. Khan’s ascent to the top of Recursion’s C-suite, after previously serving as the biotech’s chief R&D officer and chief commercial officer for a year and a half, came on the heels of a 20% layoff round that followed a pipeline cull and merger with fellow AI outfit Exscientia.
This included saying goodbye to a phase 2 candidate for a rare central nervous system disease that showed disappointing efficacy.
Khan led the pipeline shift and the Exscientia merger, she told Fierce, and the decisions were guided by data. She wanted the pipeline to be focused on programs with clear potential, and Exscientia brought an expertise that is close to home for her: chemistry.
“Ideas from biology are great, but if you can’t make molecules, then where are my products?” Khan, whose Ph.D. is in organic and computational chemistry, told Fierce.
Her scientific background is one reason why she thinks she’s the right person to lead Recursion into its next chapter, she said. The other is her business acumen, built from more than six years at Johnson & Johnson, where she led portfolio strategy and AI efforts.
Khan’s predecessor, Gibson, garnered a reputation by promising that Recursion would deliver 100 new medicines in 10 years, which, suffice to say, has not come to pass. In a blog post last month, Gibson called this a “maximalist bet” that was wrong “not in direction, but in timeline.”
Khan has no interest in “relitigating the past,” she said, but does think that both ends of the AI spectrum—extreme hype and extreme skepticism—have got it wrong.
“Neither is useful,” she told Fierce. The focus should be on how AI can be used to make medicines that matter to patients, Khan said. “Let’s be humble. Drug discovery and development is probably the most humbling thing I’ve ever done.”
Hitting paydirt
Khan’s practicality masks her deep ambition. She looks at successful platform companies like Vertex, Regeneron and Alnylam and sees a path for Recursion to join them—if the company stays disciplined.
“We have to earn the right to be there,” Khan told Fierce. She thinks Recursion’s end-to-end drug development platform is the tool to make that happen.
Recursion’s foundational platform was developed by Anne Carpenter, Ph.D., of the Broad Institute of MIT and Harvard, who now serves as a scientific and technical advisor to the company. The idea was to scrutinize images of diseased and healthy cells to identify differences that could be exploited by new drugs.
Since the biotech’s 2013 founding, its data stores have grown to include 50 petabytes of information, equivalent to 50,000 terabytes. This data is fed into models to design promising candidates, and with Exscientia’s absorbed chemistry prowess, Recursion can now synthesize new compounds quickly and efficiently, Khan said.
Recursion makes about 330 compounds for every one that makes it into the clinic, she said. For the broader industry, it’s between 2,500 to almost 5,000 compounds.
“We’re doing more simulation and making less, because you only make the things you have a lot of conviction in,” Khan explained.
Under Khan, Recursion is also looking to hasten clinical development timelines by infusing AI into patient recruitment. Over the past year, she’s been leading the construction of an in-house model that can scrub a database of 300 million patients to find which sites are the best match for a given drug candidate’s trials.
“I had built a version of this at J&J as well,” Khan said. “We have our own proprietary models which tell us the factors most important to predict that a site will perform well.”
Rather than sending a protocol to an outside clinical research organization that then tells you which sites to use, Khan reasoned that Recursion could just map out where the patients are themselves.
“This is how recruitment should be done,” she said. “Making sure you go to the right patients at the right time.”
Recursion puts its platform to use not just for its own internal pipeline, but for partners as well, another key pillar of Khan’s strategy. She outlined for Fierce the progress Recursion has already made in partnerships with Sanofi and Roche’s Genentech.
With Sanofi, the goal is to hit a tough target in immunology and inflammation, with Recursion “designing the compound from start to finish” all the way to a development candidate.
With Genentech, the goal is to develop a comprehensive biological map of microglia, important brain cells implicated in numerous neurological and gastrointestinal diseases.
Related
Recursion’s comprehensive platform stands out among the surging crowd of other AI biotechs, Khan said, because the focus is on products, not new models. Recursion was the first biopharma to build a supercomputer with Nvidia, she noted; she now wants to be the first to use AI to put a new drug candidate into a partner’s pipeline.
“There are companies right now, and I won’t name names, but they just do the in silico predictions and then they give it back to the pharma partner,” Khan said. From her time at J&J, she bets that most of those pharmas then think that their own internal models are better, sparking a pointless back-and-forth.
“Who cares?” she asked. “We do the computation in silico, and we do all of the wet lab work, all of the experiments, all the way to meet a target product profile set by the partner.”
Sitting in Recursion’s New York office, far from the company’s home base and humble origins in Salt Lake City, Khan isn’t stressed about Eli Lilly’s new Nvidia supercomputer or the persistent drumbeat of new AI biotech launches. To her, there’s enough unmet medical need to go around.
“Eighty percent of the diseases that exist in the world don’t have disease-modifying medicines,” Khan said. “Who’s actually realizing value? That’s what we should be focusing on.”
