AI for Canadian charities: the playbook

If you run a Canadian charity, you have heard that AI will revolutionize your fundraising. You have also heard that AI is risky, expensive, and ethically fraught. Both stories are oversold. This playbook is the practical middle: what AI actually does for a Canadian nonprofit, where the real value is, and how to stay on the right side of Canadian privacy law while you do it.

We are writing this for the Director of Development, the Executive Director, or the marketing lead who is doing too many jobs and trying to figure out whether AI is worth the attention.

Where AI actually helps

Charity work runs on relationships, and AI does not replace relationships. What it does is take the high-volume, low-leverage work off your staff's plate so they can spend time where relationships are made. Three areas pay off:

Donor acquisition. Finding likely donors and reaching them efficiently. AI sharpens audience targeting, drafts and tests outreach at a volume a small team cannot match by hand, and helps you get more from channels like Google Ad Grants.

Donor retention. Most charities lose donors not because of a bad relationship but because of neglect, with communications that are too generic, too infrequent, or badly timed. AI lets a small team personalize and time communications at scale: the right message to the right donor segment at the moment they are most likely to give again.

Impact measurement and reporting. Federal funders and major donors want evidence of impact. AI can help synthesize program data into the reports funders require, turning days of staff time into hours, provided a human checks the output.

PIPEDA, Law 25, and donor data

This is the part that should not be an afterthought. Donor data is personal information, and Canadian privacy law governs how you handle it.

PIPEDA, the federal private-sector privacy law, requires meaningful consent for collecting and using personal information and reasonable safeguards over it. Quebec's Law 25 goes further, with stricter consent, transparency, and data-handling requirements, including obligations around automated decision-making.

Using AI does not exempt you from any of this. It raises the stakes, because AI workflows often involve sending donor data to third-party services. Practical guardrails:

• Know where donor data goes. If you use an external AI service, understand whether donor data is sent to it, where that service stores data, and whether that is compatible with your consent.
• Get meaningful consent for AI-assisted use. If donor data informs automated personalization or scoring, your privacy policy and consent should reflect that.
• Be transparent about automated processing, especially under Law 25, where there are specific obligations around significant automated decisions.
• Minimize. Do not feed an AI tool more donor data than the task requires.

The reassuring news: none of this blocks AI use. It shapes it. Charities that build these requirements in from the start use AI confidently; the ones that retrofit compliance after a problem are the cautionary tales.

Google Ad Grants: underused leverage

Google Ad Grants gives eligible nonprofits up to $10,000 per month in free Search advertising. That is a meaningful donor-acquisition budget that many charities barely use, because the program has quality requirements and managing keywords and ad copy takes expertise small teams lack.

This is a near-perfect AI use case. AI can accelerate the keyword research that finds the searches your potential donors and beneficiaries actually make, draft and test the ad variations the program rewards, and help maintain the account quality (click-through thresholds, relevant landing pages) the program requires to keep the grant.

The result is turning a dormant grant into a steady acquisition channel, often the highest-return AI move a Canadian charity can make, because the media is free and the only cost is the work of managing it well.

Bilingual content, done responsibly

For charities operating nationally or in Quebec, English-and-French content is a requirement, not a luxury. It is sometimes a legal one and always a practical one for reach. Producing everything twice has historically been a real cost for small teams.

AI changes the economics. It can produce strong first-draft French and English versions of donor communications, web content, and reports in a fraction of the time. The non-negotiable: a fluent human reviews fundraising and donor-facing content before it goes out. Fundraising appeals carry tone and nuance that machine translation gets subtly wrong, and a clumsy French appeal does more damage than no French appeal. Use AI for the draft and the speed; use a human for the trust.

A 90-day starting plan

If you are starting from zero:

1. Weeks 1 to 2: Pick one outcome, usually donor retention or Ad Grants activation, and document your current baseline (retention rate, or Ad Grants spend and conversions).
2. Weeks 3 to 4: Map your donor data and confirm your PIPEDA/Law 25 footing for any AI tool you plan to use.
3. Weeks 5 to 8: Ship one thing, such as a personalized retention email program or a managed Ad Grants account, and measure it against the baseline.
4. Weeks 9 to 12: Review, keep what worked, and decide the next use case.

Notice this is the same discipline as any AI enablement effort: one outcome, a baseline, something shippable, measured. Charities are not exempt from good practice; they just have tighter resources and stricter data rules, which makes doing it properly matter more.