A one-page AI policy for your organization
You don't need a 40-page document. You need five decisions, written down, that every employee understands.
Every organization that starts with AI reaches the moment someone asks "am I allowed to paste this?". If there is no written answer, every employee decides alone. That is exactly how customer data ends up in a public chat.
The policy we write with clients is one page. Five decisions.
1. Which tool
One approved tool for the organization, on an enterprise subscription. Not the free version. The reason: on the enterprise tier the data is not used for training, and it is in the contract.
2. For what
A short list of what is allowed: drafting emails, summarizing documents, drafts, analyzing internal files. And a short list of what is not: credit card details, payroll data, medical information, anything marked confidential.
3. Who gets access
Everyone, or by department. Our recommendation: everyone, because whoever is left out will use the free version on their phone.
4. What gets reviewed
Every output that goes to a customer or drives a business decision passes a human eye. That is not bureaucracy. It is what makes the tool trustworthy.
5. Who to ask
One name. The person who answers "am I allowed?" within an hour. If there is no such person, it is us.
Mistakes are allowed, leaks are not. Write it, post it, update it once a quarter. Do you have a written policy today?