AI policy
Charity AI policy template for nonprofit teams
Charities can use AI for productivity, but they need clear rules before staff paste sensitive data into tools. This template is a policy starting point, not legal advice or a claim that every AI product is eligible for nonprofit pricing.
Approved low-risk uses
- Draft first versions of internal notes from non-sensitive source material
- Summarise public documents, meeting agendas, and published research
- Create campaign content outlines for human review
- Turn approved policies into staff checklists and training drafts
Prohibited or high-risk uses
- Do not enter sensitive beneficiary, safeguarding, medical, legal, payroll, or casework data without explicit approval
- Do not use AI outputs as final legal, financial, safeguarding, or medical advice
- Do not publish AI-generated facts without source checking
- Do not connect AI tools to charity systems without an owner and data review
Minimum controls
- Maintain an approved-tools list with owner, data type, and renewal date
- Prefer team or enterprise controls when charity data is involved
- Require human review for public content and grant submissions
- Log material AI-assisted decisions when they affect beneficiaries, donors, staff, or trustees
Decision notes
- The OpenAI for Nonprofits benefit page should be checked before writing policy because official discounts currently apply to ChatGPT Business and ChatGPT Enterprise, not API usage.
- Microsoft, Dynamics 365, Power Apps, Notion, monday.com, Adobe, Canva, and other tools can embed AI or automation features under different licensing and data terms.
- This guide should be reviewed alongside the charity's data protection, safeguarding, and information-security policies.