AI tools are useful and easy to adopt. A practical checklist for small businesses on what to share, which settings to check and how to set team rules.

Most teams are already using AI, whether or not there is a policy for it. Someone summarizes a long email, drafts a reply or tidies up a spreadsheet. That is usually fine. The risk is in what gets pasted in along the way.

This is general guidance, not legal advice. In Canada, private businesses are covered by PIPEDA or, in British Columbia, Alberta and Quebec, by provincial privacy laws. Health information has additional rules in many provinces.

Know what kind of tool it is

Consumer AI apps, business plans and developer platforms handle data differently. Business and API products often don’t use your data to train models by default and offer stronger controls. Consumer settings vary. Read the data section of each tool’s terms before your team uses it for work.

A simple traffic-light rule for staff

  • Green, fine to use: public information, your own marketing copy, general questions, anonymized examples.
  • Amber, only in approved tools: internal documents, business plans, supplier details, anonymized client situations.
  • Red, never in general tools: client names with personal details, health information, ID and passport numbers, financial records, passwords.

Settings worth checking

  • Whether chats are used for training, and how to opt out.
  • How long conversations and uploaded files are kept.
  • Whether the tool has memory, and what it has saved.
  • Who on your team can see shared workspaces and connected apps.
  • Which connections (email, drive, calendar) the AI tool has been given.
Treat an AI tool like a new contractor: useful, capable and only shown what the job needs.

When AI is built into your own systems

If you are having AI added to a booking system, intake form or client portal, ask your developer:

  • Which provider processes the data, and in which country?
  • Is the data retained or used for training?
  • Is personal information removed or minimized before it is sent?
  • Is there a log of what the AI saw and produced?

We build with the same habits we use for any sensitive system: collect only what is needed, limit who and what can see it, and keep a record of every change.

Write it down

A one-page AI policy is enough for most small teams: which tools are approved, the traffic-light rule, and who to ask when unsure. It protects your clients and gives your staff confidence to use AI well.

Sources

From the DYOL blog. Photo by Kingsley Mkpandiok on Unsplash.

A free 20-minute conversation, then an honest assessment. You’ll know what it costs and how long it takes, in writing, at a fixed price.

Or email lloyd.delacruz@outlook.com. We reply within 24 hours.