In short
AI training for staff, sometimes called AI literacy, is about using AI tools safely and effectively in real work. It covers practical skills, such as giving clear instructions and checking results, and judgement: knowing what AI is good and bad at, when not to use it, and what your organisation’s rules are.
Start with the rules
Before anyone is taught tips and tricks, they need to know which tools are approved, what data may go into them, and who to ask when unsure. If you do not have an AI acceptable use policy, write a short one first. Training without rules tends to spread the use of whatever free tool people find first, including with customer data.
What to cover
| How the tools work | In plain terms: language models predict likely text, can sound confident and still be wrong, and do not know your business unless you give them the context. |
|---|---|
| Writing clear requests | Give the role, the task, the context, the format wanted and an example. Ask for changes in follow-up messages. |
| Checking output | Verify facts, figures, names and quotes against a source. Read every word before it goes to a customer. |
| Data protection | What counts as personal and confidential data, why it stays out of unapproved tools, and how UK GDPR applies. |
| When not to use AI | Regulated advice, sensitive conversations, decisions about individuals and anything you cannot check. |
Teach with real tasks
Generic demonstrations are forgotten quickly. Ask each team to bring two or three tasks they do every week, such as drafting quote follow-ups, summarising meeting notes or turning a call transcript into a CRM note, and work through them in the session with the approved tool. People leave with something they can use the next morning.
Separate sessions by role where you can. What helps an office manager is different from what helps a field engineer or a sales lead.
Include at least one exercise where the AI gets something wrong, such as an invented reference or a miscalculated total, and ask people to find it. Seeing a confident mistake first-hand teaches the habit of checking better than any warning on a slide.
Address worries directly. Some staff will be anxious about what AI means for their jobs. Be clear about what the business plans to use AI for and what it does not, and involve the people who do the work in deciding where it helps.
Keep it going
One session is a start. Nominate a few enthusiastic staff as go-to people, keep a shared library of requests that work well for your business, and hold a short check-in after a month to share what has worked and what has gone wrong. Update the training when tools or rules change.
Make it safe to report mistakes. If someone pastes the wrong thing into a tool, you want to hear about it quickly so you can deal with it, especially if personal data is involved.
If you trade with the EU
The EU AI Act includes a duty on providers and deployers of AI systems to take measures to ensure their staff have a sufficient level of AI literacy. Whether it applies to a UK business depends on its links to the EU market, so check the current position and take advice if you sell into the EU. Good training is sensible practice either way.
When this matters
Relevant when staff are already experimenting with AI or when you are about to roll out a new tool. Higher New Era AI consulting tiers include staff AI-literacy sessions and change-management support as part of the roadmap.
Questions
How long should AI training take?
A first session of one to two hours per team, built around real tasks, is a good start. Follow it with short check-ins rather than one long course.
Do staff need technical skills to use AI tools?
No. Most of the skill is in giving clear instructions, supplying the right context and checking the result carefully.
Should we train everyone or just a few people?
Everyone who might use AI needs the rules and the basics of checking output. Deeper training can focus on the people whose roles use it most.