What ChatGPT says about you, doctor (and how to check in 5 minutes)
In 2025, an ENT surgeon in Fez told us he had received three patients "sent by ChatGPT" over the year. Three people who put a health question to an artificial intelligence — and were given his name in the answer. The number sounds modest, but it is above all an early signal: the search for care is shifting towards AI assistants, and most Moroccan practitioners have no idea what those assistants are telling patients about them.
Run the test — right now
Open ChatGPT (or Gemini, or Perplexity) and ask three questions, the way a patient would:
- « Qui est le Dr [votre nom] à [votre ville] ? » — "Who is Dr [your name] in [your city]?"
- « Quel [votre spécialité] recommandes-tu à [votre ville] ? » — "Which [your specialty] do you recommend in [your city]?"
- The same question, in Arabic.
Three outcomes are possible. The AI describes you accurately — enjoy it, that is rare. Or it answers in generalities, cites a half-empty directory, or names your colleagues but not you. Or — the most unsettling case — it mixes you up with a namesake: we once audited a surgeon whose name, queried on its own, pointed to a real-estate agent in the same region.
How the AI builds its answer
Contrary to popular belief, ChatGPT "knows" almost nothing about local practitioners. When asked, it runs a live web search (through Bing's index; Gemini uses Google's), reads the first pages it finds, and composes its answer from whatever it can read. In other words: the AI only says what the web already says about you. If the web holds nothing but empty directory listings, the answer will be empty. If it holds a clear, accurate, well-structured page about your background and your procedures, the answer will draw on it.
There is no such thing as "ranking first on ChatGPT". An AI either cites you or it doesn't, depending on the question and the day. Be wary of anyone who guarantees you a position.
What determines whether you get cited
Four factors, in order of importance:
- Existing consistently. One name, identical across every platform. AI assistants consolidate information by "entity": three different spellings means three half-identities instead of one reputation.
- Having a reference source. One page — ideally your own website — that answers plainly: who you are, what you practise, where, and in which languages. Structured as questions and answers, it becomes the ideal raw material.
- Being present in both languages. Medical content in Arabic is nearly nonexistent online in Morocco. An AI questioned in Darija has almost nothing to read: the first practitioner who publishes becomes the answer.
- Social proof. Reviews and mentions by credible third parties weigh on what AI assistants retain — just as they weigh on the patient's decision.
And what about medical ethics?
Everything above is factual information: your actual background, your actual procedures, your exact contact details. That is precisely what Morocco's Code of Medical Ethics allows — unlike advertising, superlatives and promises, which are prohibited and, incidentally, useless to AI assistants. Communicating well and respecting the code are not at odds: they are the same work.
Where to start
With measurement. Before writing a single line, know precisely what Google and the four major AI assistants say about you today, which questions surface your name and which surface your colleagues' instead. That is exactly what our audit does — free, presented to you in person, with no commitment.