What it is
Claude is a general-purpose AI assistant made by Anthropic. You can use it on the web, in a desktop app, or on your phone. Give it a request, text, or files, and it can draft, summarize, analyze, suggest options, or ask for missing context.
Beyond basic chat, Claude includes Projects, file analysis, Artifacts, and Cowork. Projects hold standing instructions and reference documents such as your style guide, key messaging, or AI policy. File analysis lets you ask questions about uploaded PDFs, spreadsheets, and documents. Artifacts are editable working documents Claude builds alongside the conversation, from one-page handouts to simple interactive pages. Cowork handles multi-step work with your files, with you approving its actions.
What it is good at for advocacy orgs
Claude works well for many writing and reading tasks that take roughly 20 minutes to two hours, with a person reviewing the result. Some concrete patterns we see across the movement:
- Drafting and revising. Supporter emails, volunteer guides, grant report sections, board memos, social copy. A farmed animal welfare nonprofit drafts corporate outreach letters with Claude, then a campaigner edits for accuracy and tone before anything is sent.
- Summarizing. Turning a 40-page policy report into a one-page brief, or a meeting transcript into action items.
- Working with files. Upload a campaign metrics export and ask “which months had the biggest drop in open rates, and what changed?” Claude can read the spreadsheet and walk through it with you.
- Thinking out loud. Pressure-testing a campaign plan, listing risks, drafting interview questions, or laying out options before a leadership decision. Claude is a useful sparring partner; the judgment stays with you.
- Consistent team use via Projects. A conference team keeps speaker communication templates and event details in a shared project, so anyone drafting a speaker email starts from the same context.
Where it fits
Claude is the everyday generalist in this library. If a colleague is learning one AI tool, it should probably be this one (or a close equivalent).
The other tools here are more specialized. Claude Code is for technical staff who want an agent that edits files, runs scripts, and automates workflows. Gamma turns outlines into presentations. NotebookLM answers questions strictly from documents you upload, with citations. A common pattern is to develop content in Claude and hand it to one of the specialists — for example, outlining a deck in Claude and building it in Gamma.
Watch-outs
- It can be confidently wrong. Claude sometimes invents details — names, statistics, citations, legal claims — and can misread source material. Give it the relevant documents when you can, ask it to stay within them, and verify factual claims against the originals.
- Mind your data. Do not paste donor records, health information, or sensitive investigation material into any AI tool until your organization has decided what is acceptable. Start with Data security and privacy basics and Is my data used for training?.
- Keep canonical information in your usual systems. Claude conversations are temporary working context, so bring information into Claude when needed rather than relying on chat history as your source of truth.
- Keep a human in the loop. Claude drafts; a person who understands the audience and the stakes approves. This is especially true for anything donor-facing, media-facing, or legal.
Getting started
A realistic first 30 minutes:
- Create an account at claude.ai with your work email.
- Try one low-stakes, real task: paste a draft email you were going to send anyway and ask Claude to tighten it and suggest a better subject line. Compare against your own version.
- Upload a PDF you know well — a report your team wrote — and ask for a one-paragraph summary and three key takeaways. Compare the answer with the source so you see what Claude gets right and wrong.
- Create a Project. Add a short description of your organization, your audience, and any style preferences. Run the same email task inside the project and notice the difference.
- Skim your organization’s AI policy. If you do not have one, put creating an AI policy on your leadership team’s agenda.
Pricing snapshot
There is a free tier with daily usage limits, which is enough to evaluate the tool. Paid individual and team plans are billed per seat and raise limits, add team features, and include Projects and desktop capabilities; plan names and inclusions change periodically, so check current pricing on the website before budgeting.
