Scientists use Claude as a research partner to accelerate scientific discoveries

Researchers at universities and scientific centers are using Claude, Anthropic's AI model, as an active agent in advanced research processes. The initiative gained momentum after the launch of Claude for Life Sciences, which brought significant improvements in areas such as computational biology, interpretation of scientific data, and protein analysis. According to Anthropic, Claude has moved beyond a support tool to become a collaborator that works across multiple stages of scientific work, from hypothesis definition to analysis of large data volumes. In several cases, systems based on the model reduced tasks that would have taken weeks or months to just hours or minutes. Highlights include Biomni, a platform developed by Stanford University, which integrates hundreds of biomedical databases and tools into a single AI-controlled system. In initial tests, Biomni executed complex genomic analyses, such as GWAS studies, in about 20 minutes, with accuracy comparable to human experts. Another example is MozzareLLM, created at MIT, which automates interpretation of large-scale CRISPR gene editing experiments. The system replicates the reasoning of experienced researchers, identifies relevant biological patterns, and assigns confidence levels to discoveries, significantly accelerating scientific validation. At Stanford, the Lundberg Lab is testing Claude for scientific hypothesis generation, helping to decide which genes to study before running expensive experiments. The approach seeks to replace decisions based solely on human intuition with analyses grounded in molecular relationships. Despite current limitations, the cases presented point to a structural shift in how science is conducted. Anthropic believes that AI agents like Claude will become central to research flows, expanding scientific productivity and paving the way for new discoveries.