A One-Page AI Policy for Your Small Business

Once your team starts using AI, they will use it in ways you did not expect: pasting customer lists, sharing price sheets, sending AI-written replies without reading them. A short AI policy prevents most problems before they happen. It does not need to be a legal document. One clear page that everyone reads is far better than ten pages nobody reads.

What your AI policy should cover

1. Approved tools: which AI tools staff may use for work, and with which accounts
2. Never share: the data that must never go into AI tools
3. Allowed uses: what AI is encouraged for
4. Human check: who reviews AI output before it goes out
5. Honesty: no fake reviews, fake claims or impersonation
6. Mistakes: what to do if something goes wrong
7. Owner: who to ask, and when the policy is reviewed

Drafting it with AI

Write a one-page AI use policy for my [type of business] with [number]
employees in India. Simple language that every staff member can follow.
Sections:
1. Approved tools: [list the tools and accounts you allow]
2. Never share: customer personal data, payment and bank details,
   passwords and OTPs, employee personal data, supplier prices and
   margins, anything marked confidential
3. Encouraged uses: [e.g. drafting replies, summaries, SOPs, ideas]
4. Human check: AI drafts, a person reads before anything goes to
   customers, suppliers, banks or government; [role] approves [types]
5. Honesty: no fake reviews or testimonials, no false claims, no
   pretending to be someone else, no AI images of real people without
   permission
6. Mistakes: stop, tell [person] the same day, no blame for honest
   reporting
7. Owner and review date
Under 400 words. Use short bullet points.

Then edit it to match your business exactly. Remove anything that does not apply; add anything specific, such as rules for your industry.

Why each part matters

• Approved tools: business plans of AI tools often offer stronger data protection than personal free accounts. Even if you start with free tools, decide which ones are allowed.
• Never share: India's Digital Personal Data Protection Act, 2023 sets duties for how businesses handle personal data. Customers and staff trust you with their information.
• Human check: AI can be confidently wrong. One wrong price, date or promise to a customer costs more than the time AI saved.
• Honesty: fake reviews and misleading claims damage trust and can break advertising and consumer protection rules.
• Mistakes: if people fear blame, they hide problems. Quick reporting limits damage.

Rolling it out

1. Explain it in a 15-minute team meeting. Show one good example and one bad example of AI use in your business.
2. Share the page in your team group and print it near shared computers.
3. Ask each person to confirm they have read it (a simple "Yes" in the group is enough for a small team).
4. Lead by example. If the owner pastes customer lists into a chatbot, the policy is dead.

Review it regularly

AI tools change fast. Review the policy every six months, or sooner if:

• You start using a new AI tool or a paid plan
• Something goes wrong
• A customer or supplier asks how you handle their data

Final practice task

Draft your one-page AI policy with the prompt above, edit it for your business, and share it with your team this week in a short meeting. Put a reminder in your calendar to review it in six months.

Then mark this lesson complete. You have finished the course!

Where to go next

• Market your business with AI in AI for Marketing and Small Business.
• Get more from every prompt with Prompt Writing That Works.

Key takeaways

• One clear page beats a long policy nobody reads.
• Cover approved tools, data never to share, encouraged uses, human checks, honesty and mistakes.
• Roll it out in a short meeting, get confirmation, and lead by example.
• Review it every six months or when tools or risks change.