Your staff are probably already using ChatGPT, Claude or Copilot. A short written policy turns that from a risk into a habit. Here is the outline we use with clients; your attorney should review your final version.
- Purpose. Why you use AI: save time, serve customers better, stay honest.
- Approved tools. Name them, and require business accounts with training on your data turned off.
- Never enter. Social Security, card, bank and ID numbers; passwords; health information outside approved tools; anything a client told you in confidence.
- Human review. A person checks facts, prices, names and dates before anything AI wrote goes out.
- No final decisions by AI on hiring, firing, credit or a customer's eligibility for anything.
- Honesty with customers. Assistants say they are AI. No AI-written reviews or testimonials, ever (the FTC's rule took effect October 21, 2024).
- Privacy. Keep data only as long as needed; honor requests to see or delete it.
- Security. Two-factor login on every AI account; no shared passwords.
- Mistakes. If AI gives a customer wrong information, tell the owner the same day and fix it with the customer.
Want it done for you? Our AI-use policy kit includes the written policy fitted to your industry and a 30-minute staff briefing.