Welcome to the first issue of The UK Small Business Guide to AI. Every week: one job AI can take off your plate, with the exact steps. Ten minutes to read, ten minutes to do. No hype, no jargon, nothing to buy.
I've spent thirty years running small businesses and the last two building AI that answers phones for a living. Nearly everything written about AI is for people with an IT department. This is for the rest of us, me included.
THIS WEEK'S JOB: WRITE YOUR AI POLICY IN TEN MINUTES
SEE THIS
More than half of UK small businesses now use AI somewhere: emails, quotes, marketing, admin. Almost none of them say so. And the climate just changed. LinkedIn added a button for readers to flag posts as AI slop. Substack now lets readers scan any post for AI writing. Customers are getting good at spotting the robotic email, and what annoys them is rarely the AI. It's the feeling of being fooled.
Here's the trap: if a customer asks "did a machine write this quote?" and you fumble the answer, good work suddenly looks like deception. The fix costs ten minutes and turns the same question into a mark of professionalism.
TRY THIS
Jot down, honestly, where AI helps in your business. Emails, quotes, social posts, admin. Rough notes are fine.
Open your AI assistant (any of them, free versions are fine) and paste this in with your notes:
"Using these notes, write a short AI policy for my business in plain UK English, under 150 words. It should say where we use AI, where we never use it, that a person always checks the work before it reaches a customer, and that we stand behind everything we send. Friendly and confident, not legal sounding."
Read it. Fix anything that isn't true. If it says a person checks everything, make sure a person actually does.
DO THAT
Put it where customers can find it: a page on your website, a line in your email footer, the bottom of your quotes. Next time anyone asks about AI, you point at the policy instead of squirming. You've turned a gotcha into proof you run a tight ship. Revisit it every few months as your use of AI grows.
SEEN THIS WEEK
You're looking at an example right now. This newsletter is drafted with AI and then tested, edited and signed off by me, and it says so in the footer of every issue, because a guide to using AI that hid its own AI use would deserve the slop button. That's the whole point of a policy: not confessing, just never having anything to confess.
WORTH KNOWING
The AI detection tools are spreading. LinkedIn's flag button and Substack's reader scanning are live now, and AI companies have started invisibly watermarking their text. Assume anything you publish can be checked. One thing to do: if AI helps with your public content, say so somewhere, on your terms, before someone says it for you.
That's issue one. If it was useful, forward it to one business owner who keeps saying they should "look into AI".
That's it for this week.
Simon
P.S. Hit reply and tell me: do your customers know you use AI? I read every reply and the best answers shape future issues.