← All lessons / AI in practice / Lesson 11 of 12
Interactive lesson
Consistent AI: one standard, not twenty habits.
When everyone uses AI their own way, quality depends on who did the work. Learn how shared templates, a register of AI uses and clear owners make results consistent and safe.
- About 15 minutes
- Beginner friendly
- No account required
- Original diagrams
- Use shared templates and clear ownership across teams.
- Record AI uses and the checks each one needs.
Why consistency matters
Give three people the same task and the same AI tool, and you'll get three different results: different lengths, different facts emphasised, sometimes invented figures. That's a problem when the output reaches trustees, funders or the people you support.
Shared templates
Tested prompts for common tasks, with the rules built in.
A register of AI uses
One list of where AI is used, by whom, with which tool and data.
Clear owners
Each template and use has a named role who keeps it working.
Interactive
Same task, three volunteers
Three volunteers summarise the same (fictional) event feedback for trustees. First see their own prompts, then switch to a shared template.
Source feedback (fictional)
All 50 attendees completed the feedback form. Of these, 42 rated the evening good or excellent. The speakers received the most positive comments. Six attendees reported problems with step-free access. Some comments asked for better food options, different timings and another event next year. The events team agreed to confirm an accessible venue before booking the next event.
These outputs were written for the lesson. The switch compares worked examples; it is not a test proving that a template prevents errors. Check each summary against the source. No AI model is called.
Interactive
Build a shared template
Fill in the fields and copy the template into your team's shared library.
Your template
Key ideas to take away
- Template the common tasks. A tested prompt for each recurring task beats everyone writing their own.
- Build the rules into the template. “Only use the information provided” and “say when something isn't there” travel with every use.
- Keep a register of AI uses. You can't manage, check or explain what you can't see.
- Give everything an owner. Templates and tools need someone to update them when the model, the task or the rules change.
- Make the right way the easy way. Put templates where people already work, and keep the library short.
Check your understanding
3 quick questions. Try as often as you like. Your progress stays on this device; answers clear when you reload.
Want to go deeper?Methods, checklists and standards. Optional.
What goes in an AI use register
| Field | Example |
|---|---|
| Use | Summarising event feedback for trustees |
| Tool and template | OfficeAI Business · “Trustee feedback summary” v2 |
| Data involved | Anonymised survey comments |
| Human check | Events Lead checks figures before sending |
| Owner and review date | Events Lead · every six months |
| Risk level | Low |
Version your templates
Treat a template like any shared document: number versions, note what changed, and re-test it against a few saved examples when the tool or model is updated (see Lesson 07).
Training beats policing
Short, practical sessions on the templates and the reasons behind the rules do more for consistency than monitoring people.
Continue exploring
We've picked resources that explain these ideas from different angles. They are independent of Code With Impact.
- Guidance
AI Playbook for the UK Government (opens in a new tab)
Government Digital Service
Ten principles for using AI safely and effectively, written for public bodies but useful for any organisation.
- Framework
AI Risk Management Framework (opens in a new tab)
NIST
A widely used voluntary framework for governing, mapping, measuring and managing AI risk.
- Article
Charities and artificial intelligence (opens in a new tab)
Charity Commission
What trustees should consider, including an internal AI policy and keeping decisions with people.
Sources for this lesson
The feedback data, prompts, outputs and template are fictional and original to Code With Impact. The lesson draws on the frameworks listed under Continue exploring; no text or diagrams from them are reproduced.
How we reference: we link and credit, we never copy other educators' text, diagrams or code. See our full Sources & credits register.