Last week, several colleagues and I wrote about what election offices taught us in our pilot workshops on AI use policy. As we mentioned, one thing came up in almost every conversation, which is that staff need simple, clear direction as much as they need a formal policy.
So we built a tool to help with that, and it’s live now: the AI Use Dos and Don’ts generator.
Drag suggested statements into the Dos and Don’ts column. You can add your own where your office has specific rules. You add your logo and jurisdiction, then print it, save it as a PDF or copy your answers. Nothing you enter is stored.
A handful of statements are locked as permanent don’ts, and you cannot edit them away. Everything else is yours to shape.
You do not need the whole thing figured out
You do not need a finished AI strategy to give your team clear direction. Pick the statements that are true for your office today. You can revisit it before the next cycle, when the tools and the threats will have changed anyway.
The tool meets you where you already are, and it works in both directions.
If you already have policies on the books, such as a computer-use policy or data privacy rules, you are further along than you think. Much of what an AI policy needs to say is probably already written down somewhere in your office’s existing policies. Use the builder to translate that language into plain dos and don’ts your team can follow.
If you do not have anything written yet, it works the other way. The dos and don’ts you build become the seed for a fuller policy or set of policies.
A starting point, not a policy
Check every statement against your own rules, your state law, and your IT and security team before you post or share it.
None of this is legal advice. For questions about election law or public records, go to your counsel or your state election office.
And if you have questions, want to talk through how this fits your office, or just want to tell me what is missing, reach out. My email is [email protected].
Originally posted on the AI and Elections Substack, July 13, 2026. Read the original ↗