The one AI rule every business should set first
Before you roll out any AI tool, set one simple rule about data. It prevents most of the problems.
Most AI mishaps in business come down to one thing: someone put information into an AI tool that shouldn't have gone in. Before you do anything else, set one rule — and make sure everyone knows it.
The rule: never put personal, confidential or sensitive information into a public AI tool unless you know exactly where that data goes and have the right to use it that way.
That means no customer personal data, no employee records, no commercially sensitive figures, no anything-you-wouldn't-email-to-a-stranger — unless you're using a tool with proper data agreements in place.
Why it matters: many free AI tools may use what you type to improve their models, and under UK GDPR you're responsible for personal data you process. Putting a client's details into the wrong tool can be a data breach.
The fix is simple and cheap. Write a one-page acceptable-use note: which tools are approved, what may and may not go in, and who to ask if unsure. For most small organisations, that single page prevents the large majority of AI risk. Get the guardrail in place first, then let people use AI freely within it.
Executive-level expertise without executive-level overhead.
Book a discovery call