Anthropic has detailed a broader push to place Claude inside real drug-development programs. The company names six partners—BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines and The Column Group—and says it has also started internal programs, including work related to rare childhood muscular dystrophies.
Testing the model across the development pipeline
The projects extend beyond proposing molecules. Anthropic describes Claude being used for target research, protein and modality engineering, experimental planning, regulatory and operational work, and assessment of early clinical-development plans. The strategy is to identify bottlenecks by embedding the model in active programs rather than evaluating isolated benchmark tasks.
In one Generate Biomedicines case study, six scientists used Claude Code to propose shorter genetic designs for a CAR T-cell component. Anthropic says one design was about 70% shorter and produced roughly ten times as many receptors in laboratory tests, while remaining active longer and killing tumor cells more efficiently.
Useful evidence, but not clinical proof
Those results are preclinical and partner-reported. Bristol Myers Squibb is using Claude to analyze plans and estimate the probability of success for an early-stage clinical program, but that is decision support, not evidence that the model improved trial outcomes. Anthropic says the work has already informed priorities for future model improvements.
The material change is the move from generic life-sciences assistance to a portfolio of named, active development collaborations. The uncertainty is equally important: the reviewed evidence does not show an approved drug, completed clinical trial or independently replicated therapeutic benefit.