One page · fill it once
Five decisions that keep AI from costing you a client, a claim, or a channel. Write them down once, and stop making them under pressure at eleven at night.
Which tools, on which plan, for which kind of work. The tier matters: consumer plans may train on your conversations unless you opt out; business, workspace, and API tiers generally don't.
The data categories that don't touch an AI tool at any tier. Remember that uploading a document for a summary sends the whole document to a third party.
What triggers a label on published work, in your own words. The working line: disclose when AI is the subject of belief — is this person or event real? — not when it's a tool.
The line that goes in your agreements, written once and reused. Most clients don't mind that you used AI. They mind finding out from someone other than you.
The named human accountable for every claim that gets published. AI drafts the claim. A person owns it. Anything with a number, a name, or a source gets verified before it ships.
Not legal advice. Rules on synthetic media, endorsements, and AI-generated work vary by jurisdiction and change often — verify against primary sources before relying on any of it.
Your answers stay in this browser. Nothing is uploaded.