Biotech Industry · global
Basecamp Raises $140 Million to Advance AI-Designed In Vivo Cell Therapies
Moving cell engineering into the human body is the direction supported by Basecamp’s latest funding round. AI-designed gene insertion tools already have laboratory evidence, but moving from human cells to patients still presents challenges in delivery and long-term safety.
If cell therapies can modify cells directly inside the human body, they could reduce the complex steps involved in preparation outside the body. Basecamp Research is pursuing this direction: the London-based company announced on September 23, 2026, that it had completed a $140 million Series C financing round to train a new generation of EDEN biological AI models and advance AI-designed therapeutic candidates toward clinical development, initially focusing on in vivo cell therapies.
The oversubscribed round was led by venture capital firm S32, with participation from NVIDIA and Menlo Ventures’ Anthology Fund, among others. Roche Vice Chairman André Hoffmann also invested in a personal capacity. S32 General Partner and former Verily CEO Andy Conrad will join the board. The investor list was subject to a correction: BioSpace noted on September 25 that Basecamp’s original announcement had mistakenly listed Anthropic as an investor in the round; this should not be conflated with Anthology Fund’s participation.
At the core of the technology is the connection between “designing genetic instructions” and “writing those instructions into the genome.” According to the company, EDEN can design long, complex DNA sequences, which are then paired with large serine recombinases with the aim of precisely integrating those sequences into cellular genomes. If this can be achieved in vivo, the company hopes to simplify the preparation and administration of cell therapies. This remains a development objective and has not yet been demonstrated in patients.
The model’s training data come from the company’s Trillion Gene Atlas, with samples collected and sequenced through biodiversity partners in more than 30 countries. These data expose the model to a broad range of natural sequences, providing a foundation for generating new DNA and protein designs. European Biotechnology Magazine also cautioned that the database’s name expresses a vision for the project’s scale and should not be taken as evidence that one trillion genes have already been cataloged.
Existing experimental results provide preliminary support. GEN cited a preprint published on bioRxiv in January of this year, reporting that EDEN-designed recombinases had an overall functional hit rate of 63.2% in tests using different DNA prompt sequences. European Biotechnology Magazine further reported that half of the 20 proteins tested in human cells successfully inserted a CD19 chimeric antigen receptor (CAR) construct into human T cells. This supports the possibility of using the tools for cell engineering, but these percentages describe functional performance in experiments and cannot be translated into treatment success rates.
The company’s current in vivo CAR T-cell programs cover blood cancers, undisclosed autoimmune indications, and solid tumors. They remain at the lead candidate optimization or exploratory stage, and no date has been announced for the first clinical trial. The related preprint has also not yet undergone peer review. Moving from cultured human cells to in vivo treatment will require answers about how to deliver the tools to target cells, whether gene insertion is sufficiently specific, and whether the modified cells can function safely over the long term.
The funding will also support expanded partnerships with pharmaceutical companies. GEN reported that Richard Pearce, who previously led business development, strategy, and portfolio management at Biogen, became Chief Business Officer in April. The new financing provides resources for model training and therapeutic candidate development. How far this approach can go will ultimately depend on whether the activity observed in cell experiments can be translated into reproducible treatments that can be administered safely.