How to Turn Your Staff Handbook, Donor Data Rules, and Approved Tools List Into a Nonprofit AI Use Policy With NotebookLM

Many nonprofits are already using AI in small, unofficial ways, but the rules around donor data, confidential notes, public statements, and staff review often lag behind. This workflow helps you turn real internal material into a usable governance artifact instead of starting from a generic template. It is for nonprofit executives, operations leads, development leaders, board administrators, and policy owners who need something practical that staff can actually follow.

Editorial guardrail: Use AI to organize and draft. A staff member should verify every rule that affects donor data, beneficiary information, HR matters, board material, legal review, and public communications before the policy is adopted.

What you need

• Your current staff handbook or employee manual
• Any donor data, privacy, records-retention, confidentiality, or acceptable-use policies you already enforce
• A simple list of tools your staff already uses or wants to use, even if it is just a one-page note
• A Google account for NotebookLM, with desktop preferred for source upload and policy review

How to capture or gather the source material

• Export the current handbook and related policies as PDFs if they live in Word or Google Docs. One clean PDF per policy is enough.
• Create a short approved-tools note. Include the tool name, who uses it, what they use it for, and whether the tool ever touches donor, volunteer, or beneficiary information.
• If staff AI use is informal, pull together a one-page reality check from emails or a quick manager survey. List actual use cases such as grant drafting, donor emails, spreadsheet cleanup, meeting notes, and volunteer scheduling.
• Name the files clearly before upload. Good names include Staff Handbook.pdf, Donor Data Policy.pdf, Confidentiality Policy.pdf, Approved AI Tools.docx, and Current AI Use Cases.docx.

The fastest workflow

1. Create one notebook for nonprofit AI governance and upload the handbook, privacy rules, confidentiality rules, donor-data rules, and approved-tools note.
2. Ask NotebookLM for a policy brief that sorts likely AI uses into allowed, caution, prohibited, and leadership-review categories using only the uploaded sources.
3. Ask a second question for missing policy language. Focus on disclosure, human review, donor data, public-facing content, volunteer records, and board materials.
4. Turn the brief into a short draft policy with sections for purpose, scope, approved uses, prohibited uses, review expectations, and escalation paths.
5. Have a human reviewer mark what is final, what needs counsel or board review, and what should move into staff training or an internal FAQ.

Tool-specific instructions

Primary path: NotebookLM

• NotebookLM is the best fit here because the workflow starts from real source documents and you need grounded answers with inline citations back to those files.
• Keep the notebook narrow. Do not mix policy drafting with unrelated grant or donor material.
• Ask for decision categories first. That produces a stronger governance draft than asking for a full policy in the first prompt.
• Open the citations and check every statement about donor data, staff review, and confidential information before moving anything into your final policy.

Fallback options

ChatGPT fallback

• Upload the same PDFs to ChatGPT and ask for a policy outline that separates confirmed source-based rules from suggested new language.
• Use this path when you need a faster first pass and you are comfortable doing the source verification manually.

Claude fallback

• Upload the same files into a Claude Project so the documents stay available while you refine the policy language over multiple chats.
• Ask Claude to flag conflicts between the handbook, privacy rules, and the proposed AI uses before it drafts anything.

Copy and paste prompt blocks tailored to the workflow

Primary prompt

{
  "task": "Create a nonprofit AI use policy brief using only uploaded source documents.",
  "output_sections": [
    "Purpose",
    "Scope",
    "Allowed uses",
    "Use with caution",
    "Prohibited uses",
    "Human review requirements",
    "Escalation questions",
    "Missing policy language"
  ],
  "instructions": [
    "Use only the uploaded sources.",
    "Quote the source when a rule is important or uncertain.",
    "Do not invent legal requirements.",
    "Separate confirmed policy rules from suggested additions."
  ],
  "focus_areas": [
    "donor data",
    "beneficiary information",
    "board materials",
    "HR-related content",
    "public-facing communications",
    "record retention"
  ],
  "final_request": "End with a short list of policy decisions leadership still needs to make."
}

Fallback prompt

{
  "task": "Draft a short nonprofit AI use policy from the uploaded handbook and related policies.",
  "required_headings": [
    "Purpose",
    "Who it applies to",
    "Approved uses",
    "Prohibited uses",
    "Review and approval",
    "Data handling",
    "Open questions"
  ],
  "instructions": [
    "Mark any sentence that is a new suggestion rather than a confirmed source rule.",
    "Keep the writing plain and staff-friendly.",
    "Do not cite laws or requirements unless they are present in the uploaded material."
  ]
}

Quality checks

• Make sure every approved or prohibited use in the draft can be traced back to a real internal rule or a clearly labeled leadership decision.
• Check that donor data, beneficiary data, volunteer data, HR material, and board material are handled as separate risk areas instead of one generic privacy bucket.
• Verify that the policy tells staff what to do when they are unsure, not just what not to do.

Common failure modes and fixes

• The draft sounds polished but too vague: Ask the tool to rewrite the policy as concrete staff rules with examples of allowed, caution, and prohibited uses.
• The draft misses real internal constraints: Upload the actual confidentiality or donor-data policies, not a summary note, then rerun the brief.
• Reviewers disagree about scope: Add one more source that lists current use cases by team and ask for separate rules by department or task.
• The policy becomes too long to use: Keep the formal policy short and move examples, scenarios, and training notes into a companion FAQ.

Sources Checked

• Google NotebookLM Help, Learn about NotebookLM - Computer. https://support.google.com/notebooklm/answer/16164461?co=GENIE.Platform%3DDesktop&hl=en. Accessed 2026-03-27.
• Google NotebookLM Help, Add or discover new sources for your notebook - Computer. https://support.google.com/notebooklm/answer/16215270?co=GENIE.Platform%3DDesktop&hl=en. Accessed 2026-03-27.
• OpenAI Help Center, File Uploads FAQ. https://help.openai.com/en/articles/8555545-file-uploads-faq. Accessed 2026-03-27.
• Anthropic Help Center, How can I create and manage projects?. https://support.claude.com/en/articles/9519177-how-can-i-create-and-manage-projects. Accessed 2026-03-27.
• Candid, Getting started on a responsible AI use policy for nonprofits. https://candid.org/blogs/how-to-create-responsible-ai-use-policy-for-nonprofits/. Accessed 2026-03-27.
• BoardEffect, Nonprofit leaders share their thoughts on AI. https://www.boardeffect.com/blog/leaders-thoughts-ai/. Accessed 2026-03-27.

Quarterly Refresh Flag

Review this article by 2026-06-25. Re-check product features, upload flows, and nonprofit workflow references before updating or republishing.