The AI rules for your practice.
What the rules actually say, and the one habit that keeps you on the right side of all of them. Print it. Stick it near the screen you chart on.
What the APA said
The APA issued its first AI ethics guidance in June 2025. The short version: you stay responsible for anything AI drafts in your name. The tool is never on the hook. You are. A note your scribe wrote is still your note the moment it reaches the chart.
What Illinois did
Illinois banned AI from making therapeutic decisions in August 2025, at ten thousand dollars per violation. Notes and admin use stay allowed with consent. Utah and Nevada followed with their own rules, and more states are drafting. The pattern is the same everywhere: AI can help with the paperwork. It cannot decide care.
The two questions before you paste
Could this identify my client? And is this tool one my workplace actually allows? If either answer is shaky, don't paste it. Strip the identity before you ask for the help. The chat box is not a vault.
The one habit that covers all of it
The one-source rule: the judgment stays yours, and AI never decides care. Read every AI draft against what you actually saw and heard. The note can read clean and still say the wrong thing. Your job is to make sure it's also true.
The full toolkit teaches the six ways an AI note goes wrong and the check that catches each one, with practice built in. It's called Notes, Not Judgment and it costs $29 once.
See the toolkit · $29 →This is business-skills content, not legal advice. Your state board and your workplace policy come first.