Free AI acceptable use policy template for your team

Mostly scope. A generative AI policy usually covers tools that create text, images or code from prompts, such as ChatGPT, Claude, Gemini or Copilot. An AI acceptable use policy covers those plus AI features inside software you already use, meeting notetakers, browser extensions and AI that scores or recommends decisions about people. This template covers both, so you can use it as a generative AI policy template and delete the parts you don't need.

A ban is a policy too, and it needs the same parts: a definition of what is covered, a way to see whether it is followed, and a route for exceptions. Without an approved option, people who find AI useful may keep using it on personal accounts and devices, where the company has no say over the data. The template takes the other route: approve a short list of tools through company accounts, set firm rules on the data that never goes into them, and answer requests for new tools quickly. If you do ban AI tools for some roles or some data, the register's Not approved status and the policy's data classes still apply.

The template's restricted class: personal data about employees, applicants or customers; health information; financial account and government ID numbers; passwords, API keys and other credentials; privileged legal material; and any regulated data you add. Confidential information, such as customer contracts, pricing, source code and unreleased plans or financial results, goes only into tools the register approves for confidential data, through company accounts. When someone is unsure, the template tells them to treat the information as the higher class, remove names and identifying details, or ask.

The template ends with an acknowledgment block, and the rollout email asks people to acknowledge the policy by a set date. A signed or electronic acknowledgment is a record that each person received the rules, which helps when you enforce them and when you update the policy. Whether a signature is required for your locations or contracts is a question for counsel.

The template sets a review every six months or at least once a year, plus an earlier review when you approve a significant new AI tool, a law that applies to you changes, or an incident shows a gap. Each tool in the approved AI tools register also has its own owner and review date, and the workbook turns a review date red once it has passed. Announce changes before they take effect and keep earlier versions with their dates.

Not as such. SB 947, operative July 1, 2027, regulates how employers use automated decision systems in disciplinary and termination decisions: no sole reliance on one, human corroboration when a decision relies primarily on one, a description of the employee's own data on request, a written post-use notice, and no use to infer protected status, with a civil penalty of $500 per violation. A policy is the practical place to write down the human review step. The template's human review clause and its optional California note were written with SB 947 in mind; have counsel confirm the steps for your business.

No. Intelogos doesn't record what people type; it records only whether keyboard or mouse input happened. Intelogos doesn't take screenshots. It records the AI apps and websites people use like any other tool and groups them in a built-in AI time category, so on the Tools page you can filter to the AI category and to a department to see which AI tools that department uses and for how long.

No. The template is general information written by Intelogos, not by a law firm, and it has not been reviewed by an attorney. It gives you a structure and starting language to adapt with your own counsel for your states, countries, industry, contracts and AI tools. Every document in the download carries that note.