AI Tools for Small Organisations
How to use AI assistants such as ChatGPT, Claude, Gemini and Copilot safely and well for marketing and admin. You'll learn what AI can and can't do, choose and set up tools safely, write prompts that work, create marketing content, images and video, use AI for research and everyday admin, manage data protection, accuracy and risk, and build AI into how your organisation works.
Watch the 1-minute trailer
How to use AI assistants such as ChatGPT, Claude, Gemini and Copilot safely and well for marketing and admin. You'll learn what AI can and can't do, choose and set up tools safely, write prompts that work, create marketing content, images and video, use AI for research and everyday admin, manage data protection, accuracy and risk, and build AI into how your organisation works.
The course has 8 modules, each with a short video, a written lesson, an exercise, a quiz and at least one worksheet (13 worksheets in all, in Word and PDF).
Who this course is for
People in small UK businesses and charities who want practical, safe results from AI. Not technical.
Course content
Module 1: What AI Can and Can't Do
Before you choose a tool or write a single prompt, it helps to know what an AI assistant actually is, why it sometimes gets things confidently wrong, and which jobs in your organisation it's genuinely suited to. This module welcomes you to the course, explains how AI assistants work in plain English, looks at what UK courts and the vendors themselves say about made-up answers, and gives you a neutral first look at the main assistants. You'll finish with a short list of tasks worth trying first, and a list of tasks to keep away from AI.
about 65 minutes of reading, plus a 60-minute exercise (spread over a week if you keep a task log).
What you'll be able to do
- Explain in plain English how an AI assistant produces an answer, and why it has no live knowledge unless it searches the web.
- Explain why AI assistants make things up, using what the vendors say and what UK courts found in Harber v HMRC (2023) and Ayinde v Haringey (2025).
- Sort your organisation's tasks into those that suit an AI assistant, those that need extra care, and those to keep away from AI.
- Describe the main AI assistants (ChatGPT, Claude, Gemini, Microsoft Copilot and Perplexity) on the same neutral criteria, without needing to pick a favourite.
- Spot the common myths about AI tools and say what's actually true.
Module 2: Choosing and Setting Up Tools Safely
The AI assistant you pick matters less than the plan you're on and the settings you choose. This module shows you what free, paid and business plans really give you, what each of the five main vendors says about using your chats to train its models and how to switch that off, how long your chats are kept, the AI you may already be paying for in Google Workspace or Microsoft 365, the nonprofit offers available, and how to lock your accounts down. You'll finish with a side-by-side comparison of the tools you're considering and every account set up the way you want it.
about 70 minutes of reading, plus a 60-minute exercise with your AI accounts open.
What you'll be able to do
- Explain the difference between free, personal paid and business plans, and choose the right kind of plan for the way your organisation will use AI.
- Say what ChatGPT, Claude, Gemini, Copilot and Perplexity each say about using your chats for training, and switch training off on any consumer account.
- Explain how long chats are kept, what temporary or incognito chats do, and why data residency is mostly an enterprise feature.
- Find out which AI features are already included in your Google Workspace or Microsoft 365 subscription, and how a work account differs from a personal one.
- Check which nonprofit and education offers your organisation may qualify for, and cost a plan for your team.
- Secure every AI account with 2-step verification, and replace any shared logins with one account per person.
Module 3: Prompting That Works
The difference between a disappointing answer from an AI assistant and a useful one is usually the brief you gave it. This module takes the advice that OpenAI, Anthropic, Google, Microsoft and Perplexity publish about writing prompts, boils it down to a short checklist, and shows it working on real jobs from small UK organisations. You'll finish with a saved set-up in your own assistant and the start of a prompt library your whole team can use, plus a simple routine for checking what comes back.
about 75 minutes of reading, plus a 60-minute exercise in your own AI assistant.
What you'll be able to do
- Explain what the vendors' own prompting guides agree on, and what a better prompt can and can't fix.
- Write a prompt using the six parts of a good brief: who it's for and why, the role, the task, your material, an example, and the format and limits.
- Improve a first answer by following up, asking the assistant to ask you questions, and breaking bigger jobs into steps.
- Save your context so you don't retype it, using custom instructions and the project or custom assistant features in your tool (Projects, Gems, notebooks and similar).
- Build and maintain a prompt library for your organisation, with owners, review dates and notes on what must never be pasted.
- Check AI output before you use it, with a five-point routine.
Module 4: Writing Marketing Content with AI
AI assistants are very good at producing a first draft of a web page, an email, a social post or an ad in seconds. They are much less good at sounding like you, knowing what is true about your business, or knowing the UK rules on advertising, reviews and fundraising. This module shows you how to give an assistant your voice, draft the main kinds of marketing content, edit the result until it is properly yours, and run a short check before anything goes live.
about 65 minutes of reading, plus a 75-minute exercise using your own content.
What you'll be able to do
- Write a one-page brand voice guide and use it with any AI assistant so drafts sound like your organisation.
- Brief an AI assistant to draft web pages, emails, social posts and ad copy that start from your own facts.
- Edit AI drafts so they are accurate, specific and recognisably yours, and spot the signs of generic AI writing.
- Explain what Google says about AI-generated content and why mass-produced pages can count as scaled content abuse.
- Apply the ASA's CAP Code, the fake reviews ban in the Digital Markets, Competition and Consumers Act 2024 and (for charities) the Code of Fundraising Practice to AI-assisted content.
- Run a pre-publish checking routine and record who checked what.
Module 5: Images, Video and Design
AI image tools can give a small organisation a social graphic, a background or a mock-up in seconds, but they also raise questions that text tools don't: who owns the picture, whose face is in it, and whether it shows something that isn't true. This module shows you what the main image tools are good and bad at, where UK copyright law stands in October 2026, what vendors promise (and don't) about ownership and legal cover, and how to use images honestly, with consent, with good alt text and within advertising rules. It finishes with a short look at video and voice tools.
about 65 minutes of reading, plus a 60-minute exercise.
What you'll be able to do
- Choose the right kind of image tool for a job, and spot the tasks where a real photo or a designer is the better choice.
- Explain in plain terms where UK copyright law stands on AI-generated images, including section 9(3) of the Copyright, Designs and Patents Act 1988, the government's March 2026 report and the Getty Images v Stability AI case.
- Compare what each vendor's terms say about ownership of outputs and legal cover (indemnities), and what that means for your plan.
- Use photos and likenesses of real people (staff, clients, patients, beneficiaries) with consent and without misleading anyone.
- Decide when to label AI imagery, using the ASA's position, platform labels and the EU AI Act where you serve EU audiences.
- Write and check alt text with AI help, and use video and voice tools with the same care.
Module 6: Research, Data and Everyday Admin
Marketing gets most of the attention, but for many small organisations the biggest time savings from AI assistants are in the back office: looking things up, digesting long documents, tidying spreadsheets, writing up meetings and answering the inbox. This module shows you how to do each of those well, where the traps are (made-up sources, wrong formulas, people who weren't told they were being recorded, emails that try to give your assistant orders), and how to join a few steps into a simple, safe workflow you can repeat every week.
about 65 minutes of reading, plus a 65-minute exercise where you plan and test one workflow of your own.
What you'll be able to do
- Research a question with an AI assistant and check every source it gives you before you rely on it.
- Summarise long documents in a way you can check, and spot when a summary has left something out.
- Use AI assistants to clean, analyse and build formulas for spreadsheets, and test the results before you trust them.
- Use AI meeting notes and transcription lawfully: telling people first, handling personal data properly and checking the notes.
- Draft customer replies with AI while a person stays in charge of what is sent, and connect AI assistants to email and files with the least access they need.
- Explain what AI agents and prompt injection are, using the NCSC's advice, and plan a simple workflow with a human check built in.
Module 7: Data Protection, Accuracy and Risk
AI assistants are only useful to a small organisation if you can use them without leaking client details, breaking data protection law or sending out something that isn't true. This module turns the rules into habits: what UK data protection law and the ICO expect, what never goes into a chat, the extra duties some professions carry, how to stay in charge of accuracy, the new UK rules on automated decisions, the EU AI Act if you have EU customers, and when to tell people you use AI. It finishes with a light-touch risk check and a one-page AI use policy you can adopt this week.
about 75 minutes of reading, plus a 75-minute exercise.
What you'll be able to do
- Explain how UK GDPR applies when you use an AI assistant, including lawful basis, transparency, accuracy, processor contracts and when a DPIA is needed.
- Sort information into what can go into an AI assistant, what can go in only on the right plan with the right safeguards, and what never goes in.
- Recognise where professional duties (such as client confidentiality for accountants and patient confidentiality in dentistry) add to the law.
- Put a proportionate checking routine in place so that you, not the tool, stay responsible for accuracy.
- Spot when an AI use could become an automated decision under UK GDPR Articles 22A to 22D, and know when the EU AI Act's transparency rules may apply to you.
- Run a light-touch risk check on each AI use and adopt a short AI use policy for your team.
Module 8: Building It Into How You Work
Trying an AI assistant on your own is one thing. Making it a normal, safe, useful part of how your whole organisation works is another. This final module covers rolling AI out to a team, training people (including what "AI literacy" means), measuring whether it actually saves time and money, what it really costs, its environmental footprint, and how to keep up without chasing every launch. You'll finish with a 90-day plan you can start on Monday.
about 65 minutes of reading, plus a 75-minute exercise (and a few minutes each week for four weeks to keep your time log going).
What you'll be able to do
- Plan a small, safe rollout of AI assistants to a team, starting with a pilot, approved tools and individual accounts.
- Design a short staff training plan that covers safe use, checking output and your AI use policy, and explain what AI literacy means (including the EU AI Act duty if you serve EU customers).
- Measure time saved honestly, including checking time, and turn it into a value you can set against costs.
- Work out the full cost of AI tools for your organisation, including seat prices, minimum seats, VAT and currency, nonprofit offers and staff time.
- Explain the environmental footprint of everyday AI use using dated, published figures, and take proportionate steps to reduce waste.
- Set up a light routine for keeping up with changes, and write a 90-day plan with owners and dates.
Course downloads
What learners say
No reviews yet. Finish the course and be the first to review it.