Anthropic Claude

Overview

Claude is an advanced AI assistant from Anthropic capable of producing text, code, and data visualizations from a range of inputs including written prompts, uploaded files, and images. A Harvard Claude license includes access to:

  • Claude Chat: a conversational AI interface available on the web and desktop. Use it for writing and editing, research and summarization, brainstorming, data analysis, and general productivity tasks.
  • Claude Code: designed building features, fixing bugs, and navigating large projects, it runs in your terminal or integrated development environment (IDE), and gives Claude the ability to read codebases, write files, and assist with complex development workflows — building features, fixing bugs, and navigating large projects.
  • Claude Cowork: an agentic desktop tool designed to handle complex, multi-step tasks. Instead of responding to one prompt at a time, Cowork can take a goal you describe and autonomously plan and execute the steps to get there, working across your local files, folders, and applications.

Compliance with University guidelines for use of generative AI required for use.

Eligibility

Faculty, Staff, Students, and Researchers in Central Administration, FAS, Harvard College, GSAS, SEAS, GSE, HBS, HDS, HKS, HLS, HMS (Quad), HSDM (Quad), Radcliffe, and SPH. Faculty, and Staff in DCE.

Access approved and coordinated by Schools.

Privacy and security

Approved for up to Level 3 confidential data

Claude Cowork is approved for tasks involving up to Level 3 confidential data, such as drafting text or summarizing documents, provided sessions are actively monitored and guardrails are followed. However, processing Level 4 or Level 5 confidential data within active automated sessions is strictly prohibited.

Data you enter will not be used to train Anthropic's models.