Medications
AI can support your healthcare. It cannot be your prescriber.
AI is useful in healthcare. We use it in our own clinic. But a chatbot cannot examine, prescribe or take responsibility.
Well Revolution
September 20, 2026
Where AI genuinely helps your healthcare, where it cannot, and how we use it openly inside our own clinic
One line from Re: News' 2026 investigation into New Zealand's black-market peptide trade has stayed with us. Describing how buyers work out what to do with their unmarked vials of powder, one interviewee explained they ask AI chatbots, and summed it up: "It's literally like having a doctor in your phone."
We understand the appeal of that sentence better than most, because we run a medical service that uses AI every day, openly and by design. We think AI is one of the best things to happen to healthcare access in decades. And that is exactly why we want to be precise about what it is: because "like having a doctor in your phone" is the most dangerous kind of claim, the kind that is almost true.
What AI is genuinely good at
Used well, AI closes real gaps in healthcare, and pretending otherwise convinces nobody:
- It is always available. At 11pm, when the worry actually strikes, it answers.
- It explains things patiently. Medical language, test results, how a class of medicines works: AI is a superb translator, and it never makes you feel silly for asking twice.
- It helps you prepare. Organising your symptoms, your history and your questions before you talk to a clinician makes the consultation better.
- It handles the admin of care. Reminders, follow-ups, gathering the right information at the right time.
If that were the whole job of a doctor, the chatbots would have taken over already. It is not, and the missing parts are precisely the ones that keep people safe.
What the phone cannot hold
A chatbot advising someone on a black-market injection is missing everything that makes advice medical:
- It cannot see you. It does not know your blood pressure, your history, your other medicines, your kidney function, or the interactions between them. It works from what you type, and the things that make a medicine dangerous for a particular person are usually things they do not know to mention.
- It cannot verify what is in the vial. This is the black-market trap. Even careful instructions are meaningless when nobody, including the NZ Drug Foundation's testing, can confirm what the powder actually is or how strong it is. Precise guidance about an unknown substance is precision theatre.
- It cannot push back on what you want to hear. Ask a chatbot enough ways and it will usually help you get to yes. A doctor's job is partly to be the person who says no, and means it, because your safety outranks your preference.
- It takes no responsibility. When a doctor prescribes, a named, registered professional is accountable to the Medical Council and to you, backed by a complaints system and a duty of care. When a chatbot advises, nobody is accountable for what happens next. The disclaimer sees to that.
Going it alone means getting less, not more
Here is the part that matters even if nothing ever goes visibly wrong. The results that made weight-loss medicines famous were produced inside structured care: assessment, gradual dose changes, side-effect management, regular follow-up. Strip that away and you do not get the trial results at a discount. You get less: less effectiveness, worse side effects handled badly, more abandoned treatment, and no plan for keeping results once you have them.
The person injecting mystery powder with chatbot coaching is not getting more healthcare than the rest of us. They are getting less than everyone in every trial ever run on these medicines, while carrying more risk. That is the quiet cost of the doctor in your phone: not always a disaster, just consistently less than the real thing, with no one watching.
How we use AI, and where we draw the line
At Well Revolution, AI is part of the team, and we are upfront about it. Our AI care agent Charlie is often the first to respond in member chat: gathering your story, answering service questions, making sure nothing you tell us gets lost, and preparing your case so nothing is missing when it reaches clinical eyes.
Then the line. Every clinical decision, every prescription, every treatment plan is made by an NZ-registered doctor. Charlie prepares; doctors decide. AI makes our clinicians faster and our service more responsive. It does not replace the clinician, because the judgement, the authority to prescribe and the accountability cannot be delegated to software, and any service that pretends otherwise is running an experiment on its patients.
That is the standard we would suggest you hold any service to, including us. Ask two questions: "Is AI involved in my care?" and "Who makes the clinical decisions?" A trustworthy service has crisp answers to both. Evasion on either is your cue to leave.
Use both, in the right order
So use AI. Ask it to explain the medicines, help you understand the research, organise your questions. It is genuinely good at that, and being an informed patient makes every consultation better. Then take all of it to a human who is registered, accountable, and able to see the whole of you. That combination beats either one alone, and it is the only version where someone is responsible for what happens to you.
If you are weighing up weight-loss treatment, our complete guide covers how safe, legal access works: Weight-loss injections in New Zealand. And if you would rather just ask a person, that is what we are here for.
This article is general information, not personal medical advice. Your own treatment decisions always come from your prescriber.
Existing members: log in at member.wellrevolution.com and start a priority chat any time you need us.
New to Well Revolution? Start an online assessment and talk it through with an NZ-registered doctor.
References
Seeing a doctor is as easy as sending a text
Concerned, need advice, a prescription, refill or referral?
Chat with a GP now
Subscribe to our newsletter
