Free staff micro-course
Five two-minute lessons — the things every team member should know before touching AI at work. Read them, confirm at the bottom, and download a completion certificate for the staff file.
Your practice system protects patient information; the moment data leaves it, that protection is gone. So: never put patient-identifiable information into a public AI tool — no names, no clinical details, no photos, not even "just to draft something quickly." Everything else in this course is a footnote to this rule.
Go deeper: the one ruleGreen: generic tasks with no patient detail (a template letter, a training agenda). Amber: anything "de-identified" — stop and check, partial details still identify people. Red: anything about a real patient. If you'd hesitate to pin it to a public noticeboard, it doesn't go into ChatGPT.
Go deeper: Can I paste this into AI?Data escapes without anyone "using AI": browser extensions can read your practice software's screen; files saved to the desktop sync to overseas cloud backups; AI scribes store audio you've never audited. If a tool can see patient information, it belongs on the practice's AI tool register — even if nobody installed it on purpose.
Go deeper: the privacy edge mapSooner or later someone pastes the wrong thing. The practice is protected by a calm, fast first response — not by silence. Tell the nominated person straight away; acting quickly is what matters, and an honest mistake reported promptly is never the thing that gets someone in trouble.
Go deeper: first responseAI-drafted posts, web copy and review replies read beautifully and can still contain wording that breaches advertising rules for health services — testimonials, outcome promises, "best in Sydney". Nothing AI writes about the practice gets published without a named human reviewing it first.
Go deeper: the publishing checklistEducational only — this certificate records that the five lessons were read; it is not an accreditation or compliance qualification. No patient data, ever.