Key takeaways
- Google Workspace for Nonprofits includes Gemini at no cost, and Microsoft includes Copilot Chat in Microsoft 365 subscriptions.
- OpenAI and Anthropic list nonprofit business plans at $8 per user per month as of October 2026.
- OpenAI says it may train on content from Free, Plus and Pro accounts unless the user opts out.
- California Civil Code 1798.81.5 requires reasonable security for personal information, and the definition of business includes nonprofits.
- Five guardrails cover most risk: organization sign-in, no training, clear data rules, blocking unapproved AI and human review.
Can a nonprofit give staff and volunteers AI tools on a tight budget?
Yes. Business-grade AI tools for nonprofits now come free or at steep discounts from the major vendors, and those plans include admin controls and data terms that free consumer accounts lack. The real work is choosing one tool, setting a few guardrails and blocking the rest.
Staff and volunteers are already using AI to draft grant narratives, donor thank-you notes, newsletters and meeting summaries. The question for an executive director or board is whether that happens in an account the organization controls or in a dozen personal accounts nobody can see.
Why not let people use free AI accounts?
Because free consumer accounts are built for individuals, under terms the organization never agreed to. OpenAI's help center says that for Free, Plus and Pro plans, "we may use your content to train our models" unless the user opts out, while its business offerings are not used for training by default. A volunteer who pastes a client intake note or a donor list into a personal account has moved that data outside the organization, and the organization usually has no record it happened.
California law still expects security for that data. Civil Code section 1798.81.5 requires a business that owns, licenses or maintains personal information about California residents to implement and maintain reasonable security procedures and practices. Section 1798.80 defines a business to include organizations whether or not they operate at a profit. The statute does not define reasonable security, which is why documented, enforced controls carry weight if something goes wrong.
What free AI tools can nonprofits get?
Google and Microsoft include AI chat at no extra cost for eligible nonprofits, and OpenAI and Anthropic offer discounted business plans. As of October 2026, the vendors list the following. Confirm eligibility and current pricing before you buy, because these offers change.
- Google: Workspace for Nonprofits includes access to Gemini and Gemini Notebook at no cost for eligible organizations.
- Microsoft: Microsoft's nonprofit offers page says Copilot Chat is included in Microsoft 365 subscriptions at no extra charge, lists Microsoft 365 Business Premium at $5.50 per user per month paid yearly, and lists a discounted Copilot add-on at $25.50 per user per month.
- OpenAI: OpenAI for Nonprofits prices ChatGPT Business standard seats at $8 per user per month billed annually, or $10 billed monthly, and offers up to 75 percent off ChatGPT Enterprise through its sales team.
- Anthropic: Claude for Nonprofits lists the Team plan at $8 per user per month for registered 501(c)(3) organizations, with a two-seat minimum.
The best choice is usually the AI built into the platform you already run, because sign-in, multifactor authentication, retention and admin controls are already in place. One approved tool is easier to support, train on and monitor than three.
What guardrails should come with the tool?
Five guardrails cover most of the risk, and none of them needs a large IT budget.
- Organization sign-in only. Everyone uses an organization account with multifactor authentication. Volunteers get accounts that expire when their role or project ends.
- No training, defined retention. Confirm the plan's training default in the vendor's terms and set chat and file retention in the admin console.
- Data rules people can remember. Publish a short list of what never goes into any AI tool: Social Security numbers, case notes, health information, donor payment details and anything a funder or client agreement restricts. A donor data cleanup shrinks what is at risk in the first place.
- Block or warn on unapproved AI. Use web filtering on organization devices to steer people to the approved tool, and turn on new-app alerts so you know when someone signs up for a different AI service with an organization email address.
- Human review. A staff member checks AI output before it goes to a donor, client, funder or board member.
How do you handle volunteers on personal devices?
Keep the work inside the organization's accounts, even when the device is personal. Volunteers sign into the approved AI tool with their organization account, ideally in a separate browser profile, and keep files in organization storage. When a volunteer leaves, disabling the account removes access to the AI history and the files at the same time.
For volunteers who handle client or health information, limit access to the systems that hold it, and consider whether that work should happen only on organization devices.
Which free frameworks help a nonprofit govern AI?
Two free resources give a lean organization a structure without consultants. The CIS Controls v8.1 AI and LLM Companion Guide, released in April 2026, maps familiar safeguards such as asset inventory, secure configuration, identity and logging to AI tools. NIST AI 600-1, the Generative AI Profile, lists risks that generative AI creates or worsens, including data privacy and information security, with suggested actions.
For most nonprofits, the practical version is four questions the board can understand: which AI tools are in use, what data may go into them, who has access, and how the organization would know if something went wrong.
What should a nonprofit do this quarter?
Small, steady steps work better than a big rollout.
- Ask staff and volunteers which AI tools they use now, without blame.
- Pick one tool from the platform you already have and apply for the nonprofit offer.
- Turn off training where the plan allows it, set retention and require multifactor authentication, one of the free security wins every nonprofit can put in place.
- Write a one-page AI policy and have the executive director or board approve it.
- Turn on web filtering and new-app alerts for AI services.
- Check in after 90 days on what people use the tool for and what needs adjusting.
How s90 helps
s90 helps nonprofits put AI to work within their budgets. We set up nonprofit licensing, configure AI guardrails on the approved tool, add web filtering for unapproved AI apps and turn on new-app alerts so leadership knows when something new appears. If you want a plan sized to your staff, volunteers and budget, we are glad to talk.


