AI at work

Using ChatGPT safely at work: a plain-English guide for teams

AI tools like ChatGPT can save your team hours every week. A few simple rules help you get the benefits without putting customer data, or your reputation, at risk.

Asim Hafeez · 3 min read
In short

Keep personal and confidential information out of AI tools, check everything before you use it, agree which tools and settings your team may use, be open with customers about AI, and keep a person responsible. Write these rules on one page so everyone follows the same ones.

In many businesses, staff are already using AI tools, sometimes without anyone having agreed how. Banning AI rarely works and means missing out on real time savings. A better approach is to agree a few simple rules together.

Keep personal and confidential information out

Don’t paste in customers’ personal details, bank details, passwords, health information, staff records or confidential business documents. Data protection law (UK GDPR) still applies when you share personal information with an AI tool.

Instead, use placeholders. The AI doesn’t need real names to write a good reply.

Try this
“Write a friendly reply to a customer called [Name] whose delivery of [product] is running two days late. Apologise and offer a new delivery date of [date].”

Check everything before you use it

AI can sound very confident and still be wrong. It can get facts, figures, names and dates wrong, or invent them. Always read what it writes before you send or publish it, and double-check anything important against a reliable source.

Choose your tools and settings carefully

Agree which AI tools your team may use for work, rather than everyone picking their own. Check each tool’s settings and terms: some use your conversations to improve their models unless you switch that off.

Business or team plans often come with stronger privacy controls. If your team uses AI every day, it’s worth comparing them.

Be honest about AI use

If a chatbot talks to your customers, say clearly that it’s an AI assistant. Don’t present AI-written content as expert advice, and keep a real person responsible for anything that goes out in your business’s name.

AI tools can produce text or images that closely resemble existing work. Don’t publish AI content as your own original work without reviewing it, and don’t paste other people’s confidential documents into AI tools.

Write it down on one page

Put your rules on a single page: which tools are allowed, what must never be pasted in, who to ask when unsure, and the golden rule of checking before sending. Share it with the whole team and revisit it every few months, as AI tools change quickly.

Try this
“Write a one-page AI use policy for a team of 8 at a small [type of business] in the UK. Use plain English, with a short list of do’s and don’ts.”
Not legal advice

This guide covers good everyday habits, not legal advice. If your business handles sensitive information, such as health or financial data, check your obligations with a qualified adviser. The ICO, the UK’s data protection regulator, also publishes guidance on AI and data protection.

Asim Hafeez Asim Hafeez Founder of Ash Digital & AI Solutions. 26+ years making technology work for businesses, explained in plain English. More about Asim
Discovery call

Want your team using AI confidently?

My one-to-one AI training sessions are built around your real tasks, and include a safe-use checklist your team can keep.