AI policy basics for small firms
An AI policy sounds like a big-company document. It isn't. For a small firm it's two pages that answer the questions your team, your clients and your insurer are already asking.
Why you need one now
- Your team is already using AI. A policy turns guesswork into clear rules.
- Clients are asking. Larger clients increasingly ask suppliers how they use AI with their data.
- Insurers and regulators are paying attention. Being able to show a written policy, and that people follow it, puts you in a much better position if something goes wrong.
What it should cover
1. Approved tools. Which AI tools can be used, on which plans, and who decides if a new one is added.
2. Data rules. What must never be entered, and what can be entered with care. Keep it specific to your work.
3. Supervision. Who checks AI output before it reaches a client, and how that check is recorded where it matters.
4. Transparency. What you tell clients about your use of AI, and where (privacy notice, engagement letters).
5. Incidents. What counts as a problem (wrong data entered, an inaccurate draft sent), who to tell, and what happens next.
6. Ownership and review. One named person owns the policy. It's reviewed at least every six months, because the tools change that fast.
Keep it usable
A policy nobody reads protects nobody. Write it in plain language, keep it to two pages, and walk the team through it in 30 minutes with real examples from your own work.
Start from a template
You don't need to write it from scratch. Our free Safe AI Policy template is built for small professional firms and covers every section above. Adapt it, have it checked if you're regulated, and put it in front of the team.