# How do I use AI sustainably in a charity?

> How a charity can use AI in line with its values: pick tasks that pay back, use light tools, set a short policy, ask suppliers and talk about it honestly.

Source: https://greenai.guide/how-do-i/use-ai-sustainably-in-a-charity/  
Updated: 2026-11-19

## In short

Use AI where it frees staff and volunteer time for your mission, choose the lightest tool that does the job, and avoid heavy uses such as bulk image generation. Write a short policy that covers environmental impact alongside data protection, ask suppliers simple questions about energy and location, and be honest with supporters about how you use it.

## Steps

1. List the tasks where AI saves real time for your mission and drop the ones that are novelty.
2. Use the simplest tool that works, and prefer short text outputs over images or video.
3. Write a one-page AI policy covering data protection, quality checks and environmental impact.
4. Ask your software suppliers which AI features are on by default and where they run.
5. Tell supporters plainly how and why you use AI, without green claims you cannot evidence.

**Use AI where it gives time back to your mission, choose light tools, and be open about it.** A charity's AI
footprint is usually small; the bigger risks are wasted money, data mistakes and a gap between your values and your
practice.

AI is already routine in the sector. The Charity Digital Skills Report 2026 found that 79% of UK charities
routinely use AI, while only 36% of small charities had an AI policy. Interim findings earlier in 2026 found 39% of
charities were worried about AI's impact on the environment. The figures are from the UK, but the questions apply to
non-profits everywhere.

## Where is AI worth it?

Drafting and summarising, translating, sorting enquiries and making documents accessible often pay back quickly.
Be more careful with tasks involving beneficiaries' personal data, and with generating images: in tests by Luccioni,
Jernite and Strubell, image generation used on average about 2.9 kWh per 1,000 images, compared with 0.002 kWh per
1,000 text classifications.

## What should our policy say?

Keep it to a page: which tools are approved, what data must never be pasted in, who checks outputs, and a line on
using the lightest suitable tool. See [write a green AI policy](/how-do-i/write-a-green-ai-policy/).

## Common mistakes

- **Building your own AI** when an existing tool would do. It is rarely cheaper or greener for a small charity.
- **Leaving features on** in office software that nobody uses.
- **Claiming to be "green"** about AI without evidence; see
  [honest claims](/how-do-i/make-honest-claims-about-green-ai/).

Our [guide for charities](/guides/charities/) covers free tools, supplier questions and beneficiary data.

## Sources

- [Trust issues rise as charities' AI adoption continues (Charity Digital Skills Report 2026)](https://www.civilsociety.co.uk/news/trust-issues-rise-as-charities-ai-adoption-continues.html), Civil Society (2026)
- [AI adoption among charities shows no sign of slowing down (interim findings, Charity Digital Skills Report 2026)](https://www.civilsociety.co.uk/voices/zoe-amar-ai-adoption-among-charities-shows-no-sign-of-slowing-down.html), Civil Society (2026)
- [Power Hungry Processing: Watts Driving the Cost of AI Deployment?](https://arxiv.org/abs/2311.16863), Luccioni, Jernite and Strubell, ACM FAccT (2024)
