AI Training · Free · Independent
Free, practical AI training for people who work in charities.
Learn to use AI well and safely in the work in front of you (bids, thank-yous, reports, board papers) in short, clear lessons built for UK charities. Independent and impartial: no AI company sponsors it, so it teaches what helps you.
Lesson 1 is a short film, no account needed. After that, a free account with just your email: no password.
- Free for everyoneNo paid version, no catch. Lesson 1 needs no account at all.
- IndependentNot sponsored by any AI company. No product to sell you.
- Built for UK charitiesReal charity work, UK GDPR and the rules your trustees care about.
- 80 short lessonsThree levels, about 12 hours in all, 5 to 15 minutes at a time.
What you will be able to do
Nothing assumes you have used AI before. By the end you can:
- Get a usable answer from AI on a real piece of work.
- Know what is safe to share with an AI tool, and what never to paste in.
- Check what AI gives you before anyone relies on it.
- Explain how your charity uses AI to trustees, colleagues and supporters.
- Understand how AI works, where your data goes, its environmental cost and its bias.
- Build your own way of working with AI around the job you do.
Independent and impartial
Most free AI training is made by the companies selling AI. This course is not. No AI provider sponsors it, pays for it or has any say in what it teaches. It is written for people in charities, by someone from the sector.
So it does not push one tool. The skills work in whichever one your charity uses, and it is straight about each: where it helps, where it falls short, where your data goes and what it costs, including the cost to the planet.
- Microsoft Copilot
- ChatGPT
- Gemini
- Claude
- Canva AI
- Grammarly
- Meeting notetakers
- Perplexity
In the interest of openness: the AI tools elsewhere on PurposeKit run on Claude, which PurposeKit pays Anthropic for as an ordinary customer. That is a choice of supplier. Nobody sponsors the course or shapes it.
How it works
Watch the short introduction
Lesson 1 is a film of about a minute and a half: why AI matters for charities, its risks, and how the three levels of the course work. No account needed.
Create a free account to carry on
Enter your email and we send you a code, no password. The account opens lesson 2 onwards and saves your progress on any device.
Move up through the levels
Starting out, then Understanding, then Personal mastery. All three levels are open: use AI well, understand how it works, then build your own way of working.
One path, three levels
A single straight line from your first conversation to building your own way of working. By the end of Level 1 you can get a usable answer from AI, know what is safe to share, check what it gives you, and explain how your charity uses it.
- Level 1: Starting outWhy it matters, your first real conversations, quick wins you can use straight away, and how to stay safe, legal and trusted.4 modules · 27 lessons · open now
- Level 2: UnderstandingOpen the mystery box: how AI works, where your data goes, its environmental impact, bias, and how others (and your tools) use it.5 modules · 28 lessons · open now
- Level 3: Personal masterySet up an AI workflow around how you work, then go deep on Scale, Lift and Unlock.4 modules · 25 lessons · open now
Inside Level 1: Starting out
IntroductionWhy this matters, what AI is, where it adds value, and how to approach it. · 7 lessons
- Why this course — A short film: the problems charities take on, where AI can help and harm, and how the three levels of the course get you ready.
- What do we mean by AI? — The mystery box: whatever goes in, something comes out. It covers a lot more than chatbots.
- How AI adds value: Scale, Lift, Unlock — Where AI does not help, and the three ways it does: Scale, Lift and Unlock.
- The risks and negative impacts of AI — An overview of the main risks, and how to judge whether the good outweighs the harm.
- AI affects your charity, whether you use it or not — Whether or not you use AI, your supporters, beneficiaries, funders, staff and competitors will. How to prepare.
- Three habits: curious, cautious, take action — A simple way to behave with AI, and one small step to take this week.
- Recap: the Introduction — A quick look back at the module, and three questions to see what stuck.
Getting started: your first conversationsPick a tool you are allowed to use, use it safely, have a real conversation, then learn how to get much better results. · 9 lessons
- Choosing a tool you’re allowed to use — Find out which tool your organisation has approved, or what to do when there isn’t one.
- Using it safely: the newspaper test — One simple rule for what to type into an AI tool: assume nothing is private.
- Get to your first message — Short, step-by-step help to get to the point of typing your first message in your chosen tool.
- Your first conversation: the test drive — Give the AI a role and talk to it for 15 minutes, or five if that is all you have. Do this lesson in your AI tool.
- What did you notice? And how to spot a wrong answer — Reflect on how the conversation felt, then learn the features of an AI answer: where it goes wrong, what you can do, and what is not fixed.
- Getting better outputs: the two levers — Back to the mystery box: you have two ways to change what comes out, and one is far more powerful.
- Better inputs: context, memory and prompting — Three ways to improve what goes in: give it documents, let it remember you, and ask well.
- Round two: do better inputs give better outputs? — Go back to your AI for another 15 minutes with better inputs, and see whether the answers improve.
- Recap: your first conversations — A quick look back at the module, three questions to see what stuck, and a check on the step you promised yourself.
Quick value-add winsScale, Lift and Unlock in practice: things you can try this week, plus voice, images and the first look at agents. · 6 lessons
- Scale: summarise and write faster — The easiest place to start: let AI do the first draft and the first read, so you spend your time on the judgement.
- Lift: better work and new insight — Use AI as a critical friend and a second pair of eyes, as well as a writer.
- Unlock: “hire” someone you couldn’t afford — Ask who you would have to pay to do this, then try a first draft with AI. Plus: learning adjacent skills.
- Voice, images, video and files — AI can work with more than text. Meet the multimodal side: what goes in and what comes out.
- Agents and skills: AI that does things — A first look at the next step beyond chat: tools that act, and reusable assistants you can set up.
- Recap: quick value-add wins — A quick look back at the module, and three questions to see what stuck.
Compliant, safe and trustedThe rules that apply, what supporters expect, how to be open about AI, and why every charity needs a policy. · 5 lessons
- The rules: law, guidance and your own policy — What the law and regulators say today, what may be coming, and where your own policy fits.
- Public perception: what supporters will accept — Law sets the minimum. What supporters expect, and what keeps their trust, is a separate question.
- Best practice: being open about AI — Go beyond what the law requires: say when AI is involved, and explain it in plain English.
- Your AI policy: you must have one — Why every charity needs an AI policy, what goes in it, and how to draft a first version in minutes.
- Recap: compliant, safe and trusted — A quick look back at the module and the whole of Level 1, and three questions to see what stuck.
Inside Level 2: Understanding
How AI worksA plain-English look inside the box, from a film-night decision to billions of weights, and where your data really goes. · 7 lessons
- How a box of numbers picks your evening — Weights, in a decision you make every week: what to do tonight.
- From twelve numbers to billions — The same idea with vastly more inputs, and how that becomes a large language model.
- Where your words go when you press send — Servers, who can see it, how long it is kept, and what to check.
- How a model learns, and what happens to your chats — How they learn in the first place, and what happens to your chats.
- Whose data was it? Training data and copyright — What models learned from, and the argument about whether anyone had the right.
- A glossary of geeky terms — The words you will hear, in plain English, grouped so they stick.
- Recap: how AI works — A quick look back at the module, and a few questions to see what stuck.
Environmental impactWhat the impact is, how to see it in proportion, and what you can do about it. · 5 lessons
- What does AI cost the planet? — Training versus everyday use, power and water, and why the local effects matter most.
- One chat, one kettle: putting it in proportion — How AI compares with other things, and why proportion is not an excuse.
- Lighter ways to get the same job done — The right model for the task, and when not to use AI at all.
- Use your voice as a buyer — Who is working on this, and practical ways to push for change.
- Recap: environmental impact — A quick look back at the module, and a few questions to see what stuck.
BiasBias in us, in AI, and what to do about both. · 4 lessons
- Same CV, different name: our own bias — How quick judgements about people shape who gets a chance, and how to catch them.
- How a machine learns to be unfair — Data, design and deployment, with real examples.
- The swap test, and other bias checks — Practical habits for prompts, review and decisions about people.
- Recap: bias — A quick look back at the module, and a few questions to see what stuck.
How others are using AIEveryone around you is using AI differently. See the whole picture. · 6 lessons
- What people ask AI at home — How people use AI tools in their own lives, and why it matters at work.
- The AI nobody chose — The AI people meet without choosing it: chatbots, recommendations and automated decisions.
- What your colleagues are doing quietly — What your colleagues are doing, officially and quietly.
- What charities are doing, in numbers — What charities are doing now, with the sector’s own numbers.
- Business, and the fake boss on the video call — What businesses are doing, what it means for charities, and a scam every charity should be ready for.
- Recap: how others are using AI — A quick look back at the module, and a few questions to see what stuck.
A deep dive into your approved toolsWhat each common tool is good at, and how to get the most from the one you have. You only need the ones you have. · 6 lessons
- Microsoft Copilot — Where it lives, what it does well, and what to watch out for.
- ChatGPT — Features, plans, and how to use them well.
- Claude — Features, plans, and how to use them well.
- Gemini — Features, plans, and how to use them well.
- AI inside other common tools — Canva and the AI features hiding in software you already use.
- Recap: your tools — A quick look back at the module, and a few questions to see what stuck.
Inside Level 3: Personal mastery
Setting up an AI workflowTurn occasional chats into a way of working that fits you. · 9 lessons
- There is no one right way — This is one way of working with AI. The point is to find yours.
- Explain your charity once — Bringing your documents, tools and knowledge together so it starts every chat already knowing you.
- An AI “second brain” — A place where your notes and knowledge build up and work for you.
- Turn the week’s mess into a list — Using AI to keep track of what needs doing.
- One home base for your AI — The place you start from each day.
- Letting AI act, one step at a time — Linking AI to your tools so it can do things, and how far to let it.
- Who drives, and who delivers? — Who is in charge of the steps, and who sees the result.
- Good habits, built in — Environmental awareness, referencing, bias checks and accessibility, built into how you work.
- Recap: your AI workflow, on one page — Pull it together into a way of working that is yours.
Scale: save time, save money, reach more peopleFind the biggest time and cost savings, and push the value out to the front line. · 4 lessons
- Where does your week go? The eight wastes — A classic checklist for spotting waste, and how to target each one with AI.
- What if it took a tenth of the time? — Stretch your thinking beyond small improvements.
- Push the value out to the front line — Get the benefit to the people who deliver your services.
- Recap: scale — A quick look back at the module, and a few questions to see what stuck.
Lift: better work, decisions and outcomesUse AI to raise the quality of what you do. · 5 lessons
- Get critical feedback, at any hour — How to ask for criticism you can use, from a reader of your choice.
- Think out loud with it — Thinking out loud with a patient partner.
- Read what you never had time to read — Reading and synthesising far more than you could alone.
- Ask your numbers some questions — Understanding your numbers, and how you compare.
- Recap: lift — A quick look back at the module, and a few questions to see what stuck.
Unlock: new skills, models and accessDo what was not possible before, and make sure the benefits reach everyone. · 7 lessons
- Start with a clear look at yourself — Your strengths, weaknesses and goals, using classic skills models.
- Discover what you don’t know you need — Start from the outcome you want and work back to the skills.
- A patient tutor: learning with AI — Personalised explanations, practice and self-testing.
- Brief it like a freelancer, review it like an editor — Getting good work from AI “hires”.
- From faster, to better, to new — How these skills ladder up to new ways of working.
- Making sure the benefits reach everyone — Who benefits, who carries the risk, and how to widen the circle.
- Recap: unlock, and the whole course — You have finished. Here is what to do next.
Where the real value is
AI creates three kinds of value for charities. Most teams stop at the first. The course is built to get you to the third.
Scale
Do more of the same, faster.
A fundraiser drafts 50 donor thank-you letters in the time it used to take to write five.
Lift
Do it better than before.
A hospice reads every comment from three years of supporter feedback and spots what really matters.
Unlock
Do what was never possible.
A small charity’s chief executive tests funding scenarios they could never have paid a consultant to model.
Who made it

Made by Steve Law, a fundraiser who has worked at the RNLI and Macmillan Cancer Support and originally trained as an engineer. PurposeKit is his way of experimenting with AI in the open: building things that help charities, while being clear about the limits.
Questions
- Who is it for?
- Anyone who works or volunteers in a UK charity: fundraisers, frontline staff, finance, comms, managers and trustees. No technical background needed, and nothing assumes you have used AI before.
- Is it really free?
- Yes. The whole course is free to anyone, and there is no paid version.
- Is it sponsored by an AI company?
- No. No AI company pays for the course or has any say in what it teaches. It is written to help people in charities. There is no product to sell.
- Which AI tool does it teach?
- Whichever one you have. The skills work in Copilot, ChatGPT, Gemini, Claude and the rest, and where the tools differ the course says so plainly.
- Can my whole team do it?
- Yes. It is free per person, so share the link: everyone makes their own free account and keeps their own progress. To talk about using it across your charity, message Steve on LinkedIn.
- Why do I need an account?
- To carry on after lesson 1 and keep your place on any device. It is free and needs only your email: no password.
- Do I need an AI tool at work?
- No. Everything in Level 1 can be done with made-up or public examples, and the first module of Getting started is about finding out what is approved where you work, and what to do if nothing is.
- What happens to what I type?
- Your progress is saved to your account, stored in the EU. Nothing you do in the course is sent to an AI. Privacy
- How long does it take?
- Most lessons take 5 to 15 minutes; the whole course is about 12 hours, including the hands-on practice. Do it in a few weeks of coffee breaks, or one level at a time.
- How do I get in touch?
- Send Steve a message on LinkedIn. That is the one place he reads and answers.
Start with one lesson
Lesson one is a short film on why this course exists and where it takes you. You will be done before the kettle boils.
PurposeKit AI Academy is made by PurposeKit, the free AI toolkit for UK charities. Try the tools · Explore the research · Read Steve’s newsletter