AI in Healthcare Community
Rules of the Road
Version 1 · 13 rules
AIiHC is a professional community for people working on AI adoption in healthcare.
It works because members share what actually happened - including what went wrong. That only stays possible if everyone follows a few clear rules. These are them.
The thirteen rules
Be a professional here, as you would anywhere else.
Speak to others as you would to a colleague in your own department. Disagreement is welcome and useful; contempt is not.
Never post patient-identifiable information.
No names, dates of birth, record numbers, images containing identifiers, or combinations of details that could identify someone. If you are describing a case, remove or change anything that could point to a real individual. When in doubt, leave it out.
Read the full Safe-Posting GuidanceOne account, one real person.
Join under your own name and keep one account. Do not create a second account to participate in spaces your main account cannot access, or to post views you would not attach your name to.
Do not post material your employer has not authorised you to share.
Contracts, procurement documents, internal incident reports, unpublished evaluation data and confidential correspondence stay where they are. You can describe what you learned without disclosing the document.
Say what you are affiliated with.
If you work for, consult for, hold equity in, or are paid by a company whose product is being discussed, say so in the post - every time, not once. This applies to everyone, not only badged vendor accounts.
AIiHC is not for selling.
No promotion, lead generation, product pitching, or recruiting into commercial funnels. Vendor-badged accounts have specific channels where governed engagement is allowed; everywhere else, the answer is no.
Nothing here is clinical, legal, or regulatory advice.
Members share experience, not advice. Do not rely on anything posted here for a decision about a specific patient, contract, or regulatory submission. Follow your own local governance.
Post it where it belongs.
Each channel has a stated purpose. Posting in the right place is how the community stays searchable and useful. If you're unsure, ask in Start Here - someone will point you to the right channel.
Be specific and honest, including about failure.
The most valuable posts here describe what actually happened, with enough detail to be useful. Accounts of things that did not work are as welcome as successes.
Represent evidence accurately.
If you cite a study, a regulation, or a vendor claim, represent it as it actually is. Say when something is your interpretation, your experience, or uncertain. Do not present AI-generated text as verified fact.
This is not the place for an emergency or a live patient-safety incident.
If you have an urgent clinical concern or a patient-safety incident, use your local escalation route - your clinical supervisor, your organisation's incident system, or your national regulator. AIiHC cannot act on it, and posting here does not create a record anywhere that matters.
Your posts stay yours.
AIiHC may summarise or reference discussions within AIiHC, with attribution, to build durable guidance for members. AIiHC will not republish your content outside the community without asking you first. Full terms on attribution, reuse and withdrawal will be published separately.
If something's wrong, report it rather than fight it.
Use the report function. A moderator will look at it. Do not escalate a dispute publicly.
Use the private concern route
Posting safely
Never post anything that could identify a patient. No images with identifiers, no dates combined with places, no details that would let someone work out who a patient was.
Don't post identifiable detail about a colleague or an institution without their permission. De-identify first. If you can't de-identify it and still make the point, don't post it.
Nothing here is clinical guidance. Everything you read is another professional's experience. Check it against your own governance before you act on it.
If something is going wrong right now
This is the wrong place. If an AI system is causing or may cause harm to a patient:
- Use your own institution's incident and escalation route, now.
- Tell the vendor and your medical device or AI safety lead, as your local rules require.
- Report to your national authority where required.
Once it's handled, the lessons are very welcome here - de-identified, and with your institution's permission.
What happens if a rule is broken
Most issues are handled with a quiet word or a small edit. Serious breaches - patient-identifiable information, undisclosed commercial promotion, or abuse of another member - are removed immediately and may end your membership. You can ask for a decision to be reviewed.
How we run this: replies, reports, moderation and appeals
Raise a concern