Insitro Offers $450k–$480k Chief Medical Officer Position

A Floor in South San Francisco, and the Bill for It

Insitro has just put down a marker on the floor of South San Francisco: more square footage, more bench space, more senior titles, and a hiring list that reads less like a backfill plan and more like a build-out.

The company is recruiting across at least seven senior or director-level slots right now, with two roles (Chief Medical Officer and Vice President of Regulatory Affairs) added in the past week alone. The full slate on Zero G Talent's Insitro hiring page includes VP of Bioassays, Cell Models, and Genomic Screening; (Senior) Director of Translational Medicine and Diagnostics; Senior Director of Project & Portfolio Management; and Senior Manager of Imaging Machine Learning.

That spread tells the story. Insitro isn't filling seats evenly across the org chart. It is loading up on the layer of scientific and clinical leadership that turns a machine-learning pipeline into an IND-filing machine. A Chief Medical Officer and a VP of Regulatory Affairs hired in the same week means drugs are moving toward the clinic, not just toward a poster. A VP of Bioassays and a senior manager of imaging machine learning, recruited alongside, mean the wet lab those leaders will command has to physically exist first.

Founder and CEO Daphne Koller has built insitro around what the company calls "Closed-Loop Physical AI": high-content biological data collected at scale and fed into ML models, then tested again in the lab (insitro.com/platform). That loop only closes if you have a lab worth closing it in. South San Francisco is where it physically exists, and the current hiring slate is the second half of that bet.

The comp is the tell. A CMO band at $450,000–$480,000 is competitive with mid-cap biotech, not with a typical seed-stage AI shop. The board's median tracked Insitro salary sits at $238,000, meaning the top of the slate runs roughly double that. That's the price of pulling senior drug developers away from established local employers. It's also the price of pulling them away from AI-native rivals, who are now watching their own pipelines get raided rather than the other way around.

Every role on the current slate is tagged South San Francisco. There is no "remote, US" hedge in the list. Insitro wants its new VP of Regulatory Affairs in the building, not on Zoom, because the bench is in the building, and someone has to walk it.

What Twenty Petabytes Buy You

The wet lab buildout is doing one job above all: feeding insitro's causal AI engine with ML-ready biology faster than the competition can. The company now operates on a corpus of more than 20 petabytes of automated cellular experiments overlaid with population-scale genetics, a data moat its peers cannot replicate by buying more GPUs. That corpus underwrites what insitro calls the Virtual Human, a genetically anchored causal AI engine that surfaces disease mechanisms rather than just optimizing molecules against them. The expanded South San Francisco lab is where those genetic hits become physical assays, and where the targets that survive those assays graduate into medicines.

The metabolic disease work shows the loop running end to end. In MASH (metabolic dysfunction-associated steatohepatitis), insitro reports its platform surfaced more than thirty times the genetic associations that clinical staging alone produced. On top of that library, a causal AI model integrates evidence across data modalities to predict which targets will clear Phase 2, the trial stage where most programs die. Insitro's own retrospective test asked the model to rank historical Phase 2 outcomes zero-shot, with no trial results in training. In the top decile of its predictions, the false-positive rate ran one in ten in metabolic disease and roughly one in five in cardiac, against a historical failure rate of fifty-seven in a hundred in both. The numbers do not prove prospective success. Insitro flags that explicitly. But they show the engine filtering targets before a molecule is ever synthesized, which is the bottleneck the rest of the industry is still trying to clear.

Neurology is where the partnership flywheel is spinning fastest. Three ALS targets that the Virtual Human identified have been nominated by Bristol Myers Squibb under an expanded collaboration, and insitro is now working with Eli Lilly and Gilead on additional programs. The ALS case is pointed: for thirty years, the field relied on a mouse carrying a mutant SOD1 gene, and compound after compound extended survival in those mice and failed in patients, because SOD1 drives about 2% of ALS cases while the TDP-43 pathology present in nearly every ALS patient went unmodeled. Insitro's pitch is that genetically anchored human data, run through a causal model, picks the right lock first.

The portfolio is moving from paper targets to clinical-stage work. Insitro's first ophthalmology candidate entered IND-enabling studies in May, derived entirely from platform insights and aimed at a genetically defined retinal disease subset. A June publication from the team described a deep learning model that redefines patient subgroups in metabolic disease using multi-modal cellular signatures, the kind of stratification tool that decides whether a Phase 2 enrolls the patients most likely to respond. The new VP of Regulatory Affairs will shepherd multiple therapeutic modalities (small molecules, biologics, oligonucleotides) across neurology, metabolism, cardiovascular, and ophthalmology, with programs advancing toward first-in-human studies.

The broader logic is that the field's failure rate is not a chemistry problem. More than nine in ten drugs that enter clinical trials fail, most often because the mechanism targeted was wrong. Targets with human genetic support are two to four times more likely to succeed in the clinic, yet only one in twenty-eight genetically supported targets has ever been pursued for an indication the genetics supports. Insitro's bet is that combining automated wet-lab biology at scale with causal modeling is the first credible way to widen that fraction, and the wet lab expansion is what makes the bet fundable.

Who's Getting Pulled Out of Their Chairs

Insitro's physical expansion has translated directly into a recruitment push that is lifting senior drug-hunters out of established Bay Area employers and onto its South San Francisco payroll. The company posted 53 open roles on its LinkedIn page as of late 2025, and Zero G Talent's own board data shows the two new postings in the past seven days: a Chief Medical Officer role paying $450,000–$480,000 and a VP of Regulatory Affairs at $313,000–$333,000. Forbes counts 275 Insitro employees as of March 2025, and a Vice President of Regulatory Affairs listing, advertised as supporting a company "rapidly becoming a clinical stage," implies Insitro is filling slots that, until recently, only Big Pharma could justify.

That pricing competes most directly with Genentech, the South San Francisco incumbent that has historically anchored the local talent market. Insitro's headquarters at 259 E Grand Avenue sits roughly five miles from Genentech's DNA Way campus, and several of the open director- and VP-level roles (regulatory affairs, translational medicine, project and portfolio management) match the job families Genentech staffs most heavily. The pressure is showing up in the wider competitive field rather than in named Genentech counter-offers: Recursion, Atomwise, and BenevolentAI are all repositioning while Insitro absorbs senior talent from the same pool.

Recursion, the Salt Lake City–based AI drug discovery peer, is defending its narrative; a Seeking Alpha bear thesis published in 2025 ran the headline "I'm Still Bearish Despite Recent Genentech News," which captures how badly Recursion needs its partnership pipeline to stay intact while insitro siphons from that same pool.

Atomwise has executed the opposite move: contracting rather than expanding. Endpoints News reported that the once high-flying Atomwise tapped a biotech veteran as CEO to lead a "pared-down company," and Pharmaphorum documented Sanofi signing a five-drug discovery deal with the slimmed Atomwise. The pattern reads as defensive: Atomwise is monetizing its platform through partnerships precisely because it cannot match Insitro's in-house hiring budgets.

BenevolentAI has gone still further, restructuring twice in roughly a year. Business Wire announced a "Major Strategic Overhaul" returning BenevolentAI to its "original mission," and MedCity News characterized it as a deliberate return to "TechBio Roots." Merck KGaA has nonetheless kept BenevolentAI in its partner rotation alongside Exscientia, suggesting the company is being repositioned as a discovery services vendor rather than as a pipeline competitor.

None of the research documents a specific Insitro hire pulled directly from Genentech by name, nor any formal Genentech counter-offer. What the evidence does show is directional: Insitro is paying clinical-stage salaries while Genentech's neighbors (Recursion, Atomwise, BenevolentAI) are cutting, restructuring, or pivoting to partnership revenue. The talent tug-of-war is real, even if the named defections stay behind closed doors.

Signals From the Partner Side

Insitro's wet lab buildout is registering on partners' balance sheets in real time. The clearest signal is Bristol Myers Squibb, which has now expanded its ALS collaboration with Insitro, most recently adding more novel targets on top of the original deal. ALS New Today and First Word Pharma both reported the latest target nomination, and BMS's own BusinessWire release confirms the structure: Insitro identifies targets with its platform, BMS holds downstream development and commercialization rights. Each expansion is an implicit vote of confidence from a top-ten pharma player that the South San Francisco lab can keep generating validated, IND-ready biology. In a field where most AI-discovery partnerships quietly die after the first milestone, repeat expansions are the rarest currency.

The strategic M&A side is moving too. Drug Target Review reported that Insitro acquired CombinAbleAI to fold small-molecule design into its "closed-loop physical AI" platform, a tightly framed bolt-on rather than a moonshot merger, suggesting management is investing the new lab capacity inward before chasing external partnerships. That sequencing matters: the lab has to produce targets before any pharma partner can pay for more of them, so building the engine first is the logical order of operations.

Competitor responses from the other side of the table confirm the pressure. Atomwise's Sanofi deal (a large enough commitment to signal that even leaner AI-discovery shops are landing multi-program pharma money rather than collapsing) and BenevolentAI's restructuring both landed in the same window. Merck KGaA hedged across the field, signing deals with both BenevolentAI and Exscientia to keep options open.

Public-market sentiment on peers is the cleanest read on whether Insitro's IPO window is opening or closing. Seeking Alpha's recent bear note on Recursion is a reminder that AI-discovery equities still trade on near-term clinical readouts more than on platform metrics. That dynamic will color any Insitro roadshow: pharma partnerships buy credibility, but the public market wants a clinical catalyst before it prices the platform. Whether Insitro's CMO and VP Regulatory postings translate into a filing window depends on the next BMS data drop and on whether the company follows Recursion's lead or holds for cleaner clinical milestones.

What This Story Leaves Out, and Why

A focused piece on Insitro's recent lab build-out and the talent war it triggered has to draw its borders early, otherwise the narrative swells into a general-purpose explainer on AI drug discovery, which would bury the actual story. This article does not retrace Daphne Koller's path from becoming Stanford's first machine learning professor in the computer science department in 1995 to founding Insitro, even though that arc is foundational context. Forbes covered it; the company's own LinkedIn page positions her as an "AI pioneer." Both anchors exist, but neither is what the wet lab expansion is about.

The piece also does not attempt a market-wide survey of AI-driven drug discovery. The current board data and the press cycle make that tempting, and a market map could easily spiral to 2,000 words of competitive taxonomy. The trade-off isn't worth it: every paragraph spent classifying the broader field is a paragraph stolen from the specific hiring data, the specific ALS collaboration milestones with Bristol Myers Squibb, and the specific talent flows that drive this story.

Financial detail beyond what is directly tied to the South San Francisco expansion is out of scope as well. The article does not attempt an Insitro valuation history, a public-market comparison to Recursion, or a full read-out of BenevolentAI's inconclusive Alzheimer's efficacy data. Those threads sit in the Counter-moves digest for a reason: they are receipts, not the spine. Pulling them forward would shift the article from "what insitro s wet lab boom is doing to the talent market" to "the state of AI biotech." One is a story with a thesis; the other is a sector report.

Two adjacent topics also sit outside the fence. The piece does not catalogue Insitro's full pipeline of neurological and metabolic disease programs beyond the Bristol Myers Squibb ALS target expansions already covered, no deep dives into specific asset mechanisms, trial readouts, or preclinical timelines, because those belong in a pipeline review. It also does not parse the technical claims of Insitro's "Closed-Loop Physical AI" platform as described on the company site, except where the platform directly justifies a hiring decision captured in the board data (the Senior Manager, Imaging Machine Learning role, for instance, is a platform-shaped hire, and saying so is fair; explaining how the imaging models actually train is not).

What is left, then, is narrower and sharper: a documented expansion in lab capacity, a documented hiring surge with named roles and salary bands, documented counter-moves from named competitors, and a documented shift in where senior drug-discovery talent is choosing to work. Everything else (Koller's academic origins, the broader AI biotech market, deep pipeline biology, and platform engineering minutiae) is bracketed out by design. Readers looking for that coverage will find it elsewhere; readers looking for what the lab expansion is actually doing to the people and the companies around it will find it here. The compensation spread — $166,000 to $330,000 with a median of $238,000 across 15 tracked roles — is the single line item that ties those threads together, and the floor in South San Francisco is where it gets spent.

Working in frontier tech? Zero G Talent tracks the openings: see every open Insitro role, browse frontier tech jobs, the companies hiring, and the people building the field.