We share how to use Claude Science, including setting up common workflows, new life sciences skills, and popular database connectors.

• Category [Enterprise AI](https://claude.com/blog/category/enterprise-ai)    
  

• Product Claude Science     
  

• Date August 18, 2026  
  

• Reading time 5 min   
  

• Share [Copy link](https://claude.com/blog/the-claude-science-product-guide#)https://claude.com/blog/the-claude-science-product-guide
  

Life sciences organizations are at a turning point in their AI adoption journeys. Deloitte's 2026 Life Sciences Outlook found that 78% of biopharma and medtech leaders expect AI to play a central role in driving major change this year, yet only 14% report full implementation of AI tools into their organization’s daily workflows. Anthropic's own internal research, drawn from interviews with researchers across chemistry, physics, biology, and computational fields, found that 91% of scientists want more AI in their research, while 79% named trust and reliability as their number-one barrier to adoption.

Claude Science (in beta) is Anthropic's answer to this problem: an AI workbench for every digital step of life science, built to run next to the scientist's data and produce results that can be traced, reproduced, and defended. It sits inside a broader Claude product family—including Claude Chat, Claude Cowork, Claude Code, Claude for Microsoft 365, the Claude Platform, and Claude Managed Agents—that life sciences organizations like Novo Nordisk, the Garvan Institute, and Benchling use for the document, regulatory, and enterprise work that surrounds the science. This guide covers which surface to reach for when, then goes deep on deploying Claude Science inside a research organization, with customer examples and a rollout roadmap.

• When to use which Claude surface for science, including Claude Science for analysis, figures, and results; Claude Cowork and Claude for Microsoft 365 document and regulatory work; and Claude Code for building production pipelines
• How Claude Science works underneath the hood, powered by a local daemon that keeps data, compute, and agents on your machines and dispatches heavy jobs to your own GPU box, SLURM cluster, or cloud account
• The five design choices that make Claude’s scientific analysis hold up under review
• A three-phase adoption roadmap—Foundation, Pilot, Scale—with what to do and what you'll see at each stage, plus the metrics that show a pilot is working
• Function and workflow use cases across discovery, analysis, and publication, from single-cell RNA-seq clustering to methods-section drafting

Get started withClaude Sciencetoday.