WhiteLab Genomics Raises $26M for AI-Designed Gene Delivery

What happened

WhiteLab Genomics, a Paris company that uses artificial intelligence to design the parts of genomic medicines that carry them into the body, has closed a $26 million (€23.2 million) Series B round.13 AVP, an investment platform with more than €2.5 billion under management, led the round. New investors Yaday Health and Blast Club joined, and existing backers Omnes Capital and Debiopharm Innovation Fund also took part.1 As part of the deal, AVP managing partner François Robinet and Daniel Teper, managing partner of Yaday Health and founder of NAYA Therapeutics, will join the board.3

The deal was one of about ten rounds in a busy October 6 funding day. Eight of those companies were AI-native or put AI at the center of their product.7 The largest check went elsewhere: chip-simulation startup Vinci raised $250 million at a $1.5 billion valuation.23 For anyone following AI drug discovery, gene editing and AI-driven protein engineering, though, WhiteLab is the more telling deal. It funds what may be the most stubborn bottleneck in genetic medicine.

The platform: AI-designed delivery, then proof in animals

WhiteLab's platform is called ALFRED, short for AI-Led Framework for Rational Exploration in Drug Design. It designs viral delivery vehicles such as adeno-associated viruses (AAVs), non-viral carriers such as lipid nanoparticles, and programmable genetic payloads. The company then tests those designs experimentally and in vivo.1 Its strongest public result came from work with the Paris Brain Institute. There, AI-engineered AAVs crossed the blood-brain barrier in animal studies with a strong brain-to-liver targeting ratio and no detectable liver signal.1 The company also says the designed capsids have high sequence novelty, which could give it distinct intellectual property and freedom to operate.3

The new money will push validation beyond AAVs into non-viral delivery and synthetic promoters, which control where and when a therapeutic gene is switched on.3 The goal is to design several parts of a genomic medicine together rather than optimizing the carrier and the cargo separately.5 The company named neurological targets including Alzheimer's, Parkinson's, ALS, lysosomal storage diseases and glioblastoma.1

One caution: these results come from animals, not people. One funding roundup made the same point directly.7 WhiteLab plans to present more in vivo data from its work with Sanofi, Cytiva, the Paris Brain Institute and Institut Imagine at the upcoming ESGCT congress.1 That will be the next real test of the platform.

Why delivery matters for CRISPR and gene editing

WhiteLab does not sell a gene editor. Its work still bears directly on CRISPR, because the main limit on in vivo editing is getting the editor to the right cells. An NIH-funded team made this point in April. Common gene-editing proteins are too large for targeted carriers like AAVs, which has largely restricted clinical use to cells edited outside the body, such as blood and bone marrow.44 That team engineered a compact Cas12f variant small enough to fit inside AAVs, with editing efficiency above 80% in human cells. Its next step is to test the enzyme packaged in AAV vectors.44

The field's progress is real but uneven. CRISPR Therapeutics has an approved, revenue-generating product in CASGEVY. Many peers, such as Prime Medicine, Metagenomi and Mammoth, remain in early clinical or preclinical stages with much smaller cash reserves.43 The field has also seen failures. Spotlight Therapeutics tried to deliver editors without viral vectors or nanoparticles, raised $66 million, and shut down after disappointing preclinical results.

In that light, WhiteLab's approach is a fairly safe bet. Better capsids, nanoparticles and promoters are useful whichever editor eventually wins. Improved AAV targeting helps compact CRISPR enzymes, epigenetic editors and conventional gene therapies alike.

The business model: partnerships first

WhiteLab's business model may matter as much as its science. Rather than taking its own drugs into clinical trials, the company has about 15 revenue-generating partnerships with biotechs and large drugmakers, including Sanofi, its CEO told Axios.6 AVP's Robinet described the appeal as getting paid for the platform and again if a partner's drug succeeds.6 One analysis described the setup as owning and licensing assets plus multiyear co-development deals with milestones and royalties.4

The reports differ on some details. Axios says the company has raised about $37 million to date with two to three years of runway.6 One data aggregator lists total funding of roughly $46.5 million.8 Another analysis puts financial visibility at more than two years.4 Founding credits also vary: some reports name three co-founders, David Del Bourgo, Lucia Cinque and Julien Cottineau,1 while another names only Del Bourgo and Cottineau.4 The differences are small, but they show that a private company's financial profile is often pieced together from partial disclosures. No outlet reported a valuation.4

Del Bourgo told Axios that WhiteLab has been approached by potential buyers and would consider an acquisition or a U.S. IPO.6 Along with plans to strengthen its Boston hub, build a West Coast presence and explore Japan and South Korea,1 this points to a company setting itself up as a partner, or a target, for larger drugmakers rather than as a standalone drug developer.

A modest round in a top-heavy market

By AI-biotech standards, $26 million is small. Isomorphic Labs and Chai Discovery together raised $2.5 billion this year, more than the roughly $2.2 billion the whole sector reportedly raised in 2025.20 Just before WhiteLab's announcement, Basecamp Research closed an oversubscribed $140 million Series C and CellCentric raised $220 million.14 Enveda raised $311 million in September, and it already has three candidates in human testing.16

One analysis argued that WhiteLab's narrower pitch may appeal to investors who are wary of broad claims that AI can transform the entire drug pipeline.4 That reading is plausible. The market appears to be splitting. A few frontier-model companies get mega-rounds, while focused, evidence-driven businesses raise smaller amounts on the strength of lab and animal data. The Series B is about 2.6 times WhiteLab's $10 million Series A from 2022 and comes four years later.4 That looks like steady growth, not a hype-driven jump in valuation.

The broader theme: proof in the physical world

The other October 6 deals reinforce the same message. RougeTx, a Leiden University Medical Center spinout, launched with a $58 million Series A to take RTX-001 toward first-in-human trials. RTX-001 is a once-daily oral drug candidate for hereditary haemorrhagic telangiectasia (HHT), a disease with no approved therapy.3336 RougeTx is not an AI company, and it has not yet dosed a patient.37 Its financing still rests on biological evidence, much like WhiteLab's.

Multiply Labs raised a $75 million Series B to automate biologics manufacturing with robotics. Its pitch is that AI is designing more therapies than the industry can actually produce.11 Its backers include AstraZeneca, which is also a customer.13 Together, these deals suggest investors increasingly see value at both ends of the AI design step: delivery and validation before it, and manufacturing after it.

There are also signs that AI protein engineering is moving into normal industry practice. Hansa Biopharma, for example, signed a partnership with Cradle for AI-driven protein design.18 WhiteLab's capsid work belongs in this category, since AAV capsids are proteins and designing them is a protein-engineering problem.

The verdict

The coverage broadly agrees that WhiteLab is valuable because it tests its AI designs in living animals instead of stopping at computer predictions.73 That consensus seems right, but it has limits. Results in animals are a long way from benefit in patients. Of all the companies discussed here, only Enveda and a few gene-editing firms have reached human testing.1643

WhiteLab's Series B is best read as a bet that delivery, more than the choice of gene editor, will decide which genomic medicines succeed. It is also a bet that a partnership model can fund that work until clinical data arrive. If the ESGCT data and new partnership deals follow, the round will look modest but well timed. If not, the company has two to three years of runway to prove its designs work outside the computer.