Recursion‑Exscientia merger eyed at $688M

AI drug discovery is the part of biotech trying to turn messy biology into something more like an engineering problem. That sounds great, but the hard part has always been fragmentation — one company has data, another has chemistry tools, another has clinical programs, and very few have all three. That is why the Recursion-Exscientia deal mattered. The real news is not a fresh 2026 development, but a completed 2024 transaction: Recursion announced the all-stock acquisition of Exscientia on August 8, 2024, and finished it on November 20, 2024. ### What actually happened? Recursion, the Salt Lake City techbio company, agreed to acquire UK-based Exscientia in a stock deal valued at about $688 million. The companies framed it as a merger, but structurally Exscientia became a wholly owned subsidiary of Recursion when the transaction closed in November 2024, and Exscientia’s ADSs stopped trading on Nasdaq. ### Why was $688 million the headline? Because that was the approximate equity value attached to Exscientia in the all-stock deal when it was announced. (ir.recursion.com) The exchange ratio was 0.7729 shares of Recursion Class A common stock for each Exscientia ordinary share, and the ownership split implied Exscientia shareholders would hold about 26% of the combined company, with legacy Recursion holders owning about 74%. ### Why did these two fit together? Recursion’s pitch has long been scale in biology — huge image-based datasets, machine learning models, automated experiments, and heavy compute. Exscientia came in from the other side of the problem, with a reputation for AI-guided medicinal chemistry and molecule design. Basically, one side was stronger at mapping disease biology and generating hypotheses, while the other was stronger at turning those hypotheses into drug-like compounds. The combined pitch was an end-to-end platform instead of a patchwork. (sec.gov) ### Why does “end-to-end” matter so much? Because drug discovery usually breaks at the handoffs. A target looks promising, but chemistry can’t make a good molecule. Or a molecule looks elegant, but the biology behind it is weak. The appeal here was vertical integration — fewer baton passes, more feedback loops, and a better chance that the data generated in one step could improve the next one. That does not guarantee drugs will work, but it does make the workflow more coherent. That’s the real strategic logic. (ir.recursion.com) ### Was this also about cash? Yes — very much. When the companies announced the combination, they said the merged business would have roughly $850 million in cash, enough to fund operations into 2027. In biotech, that matters almost as much as the software story, because AI platforms still have to survive the long, expensive wait for clinical proof. Cash buys time, and time is what most platform biotechs are really selling themselves. (ir.recursion.com) ### What was going on at Exscientia? Exscientia came into the deal after a rough stretch. It had cut programs, laid off staff, and gone through leadership upheaval. That made the merger feel less like two equally strong companies joining hands and more like consolidation in a stressed sector. Recursion got capabilities it wanted, but it also bought into a moment when AI-biotech valuations had cooled and scale looked safer than going it alone. (chemanager-online.com) ### So why are people still talking about it? Because this deal became a marker for where techbio was heading. Investors had spent years hearing that AI would speed up drug discovery, but the field still lacked many clear clinical wins. So the Recursion-Exscientia combination landed as a different kind of bet — if one AI company is not enough, maybe a bigger, more integrated AI company has a better shot. That is less a victory lap than an admission about how hard the problem still is. (fiercebiotech.com) ### Bottom line The catch is simple: the merger solved a platform problem, not the underlying biology problem. Recursion did close the $688 million Exscientia deal, and it did create a larger AI-drug-discovery company. But the whole thesis still lives or dies on something much less glamorous — whether the combined machine can turn better software and better chemistry into actual medicines. (ir.recursion.com)