How We Use AI
The short version
AI drafts. You decide.
We use AI tools to help us draft, edit and organize work for your business. A person on our team checks what AI produces for the work we deliver, and nothing goes public until you approve it.
Tools we use
On this page, "AI" means generative AI: tools that write text or create or change images, audio, video or code.
For now, we don't put your material into any AI tool. Before we do, we'll pay for a business AI plan for our company, check what that vendor says about training its AI on customer data, and list the plan here. We never use free or personal AI accounts for your work.
In lessons, you work in your own accounts, including your own AI account, and you decide what goes in.
When we list a plan here, the vendor's statement will be theirs, not our guarantee. "Not used for training" is not the same as "never stored." Vendors keep some data for a limited time, and longer if a law or court requires it. We'll delete our AI chats about your project within 30 days after it ends, unless the law or a dispute requires us to keep them, or you ask us to keep them. We'll recheck these terms at least every six months and update this page.
What AI does in our work
We may use AI to:
- Draft website text, product descriptions, social posts, emails and training handouts. When AI writes product titles or descriptions for Google Shopping, we mark them the way Google requires.
- Suggest ideas, outlines and summaries.
- Draft code and settings for websites and online stores. A person tests them before use.
- Clean up photos, for example by fixing the lighting or removing a background. This never changes how a product looks.
- Make images, but only when you ask. We label them as AI-made and make sure they carry the labels Google Shopping requires.
What people do
- A person on our team reviews everything AI produces for the work we deliver to you, before you see it.
- When your team uses an AI tool we set up or teach, including in our workshops and training, checking what it produces is up to your team; we teach you how.
- We check names, prices, dates and claims. AI can still be wrong, so please check facts before you rely on them.
- You approve everything before it is published.
- When we deliver your project, we tell you which parts used AI.
What we never do
- Paste private information about your customers, health information, card numbers or passwords into AI tools.
- Opt in to anything that lets an AI vendor train on your material, or send a vendor feedback (such as a thumbs up or down) on work about your project.
- Write fake reviews or testimonials, with AI or without it.
- Present AI-made images as real photos of your products.
- Claim that AI did something it did not do.
- Give legal, tax, accounting, investment, securities, financial-planning, insurance, medical or pharmacy advice, or make HIPAA compliance determinations or sign-offs.
- Take on work that touches protected health information, or work for the Town of Lancaster.
Your choices
- You can ask us not to use generative AI on your project, or on any part of it. We will do that work without it. It may take longer or cost more, and we will tell you before we start. Some cameras, apps and platforms have built-in AI features we can't turn off; we'll tell you if one affects your project.
- You can ask us at any time which AI tools we used on your project, and how.
The two habits we teach your team
In our free 90-minute AI lunch workshop and in staff training, we teach two habits:
- Read before you publish (it's sometimes confidently wrong). Check four things: Names · Prices · Dates · Claims.
- Don't paste anything you wouldn't hand a stranger (customer or patient info, contracts). The billboard test: if you wouldn't put it on a billboard on Main Street, don't paste it in.
We also ask workshop attendees not to put customer names, reviews or messages into free AI apps.
Questions? Email us at don@smallbusinessmagnet.com.
Last updated 2026-10-26.