Your Firm Needs an AI Policy. A One-Pager Beats a Ban.

The instinct at a lot of small firms is to ban AI and move on. It feels safe. It is not. Your staff are already using AI, on their phones if not on your systems, and a flat ban just means they do it without rules. The North Carolina Bar made this point plainly at the start of 2026: a realistic policy beats a ban you cannot enforce.

The good news is that a useful AI policy for a small firm fits on one page. You do not need a corporate manual. You need clear answers to a few questions.

Start with a traffic-light system. Green light: tasks anyone can use approved AI tools for, like drafting internal notes or summarizing a public document. Yellow light: tasks that need a supervisor's sign-off. Red light: things that are never allowed, and the top item here is pasting confidential client information into a free, public AI tool. That one line prevents the most common and most serious mistake.

Next, name your approved tools. List the specific AI products your firm has vetted and pays for, the ones with privacy terms that protect client data. Anything not on the list is off limits for client work. This turns a vague worry into a simple yes-or-no for your staff.

Then, set the review rule. Every piece of AI output that goes to a client, a court, or another party gets reviewed by a person first. This is the rule that keeps you off the list of firms sanctioned for filing AI-invented case citations. Texas and others now require this human check, and it is the cheapest insurance you will ever buy.

Finally, decide on client disclosure. Some states, Florida among them, require telling clients when AI affects their bill. Even where it is not required, a line in your engagement letter noting that you use AI tools with human oversight builds trust rather than eroding it.

That is the whole policy: traffic lights, approved tools, mandatory review, client disclosure. Write it down, share it with your team, and update it when your tools change. A firm with a clear one-page policy can use AI confidently. A firm with a ban is just hoping nobody breaks a rule that was never written down.