28% of Instagram's Self-Described Plastic Surgeons Were Not Board-Certified. No One Has Checked What AI Does With That.

325 of 1,141 Instagram "plastic surgeons" were not board-certified

The strongest evidence that the label "plastic surgeon" is unreliable has nothing to do with AI. In January 2025, PRS Global Open reviewed 1,399 US physicians marketing themselves on Instagram. Of those, 1,141 explicitly listed themselves as plastic surgeons. 325 of them, 28%, were not certified by the American Board of Plastic Surgery; 251, 22%, had no plastic surgery training at all (PRS Global Open via PubMed). It is peer-reviewed and the authors declare no financial interest.

Read that as a statement about sources, not about surgeons. Instagram bios, practice sites and directory profiles are the crawlable public text a model reads when a patient asks who to trust with their face, and in this sample roughly a quarter of it described a credential that was not there. Nothing in "Dr. A is a leading plastic surgeon in [city]" separates a physician who completed an ABPS-accredited residency from one who typed the words. The patient cannot see the difference either. Certification is a fact; "plastic surgeon" on a profile is a claim. On the shortlist an assistant produces, they arrive in the same font.

No one has measured whether ChatGPT tells a plastic surgeon from a cosmetic surgeon

This has to be said plainly, because the rest of the article is worthless without it. No published study measures whether ChatGPT, Gemini or Perplexity recommend non-board-certified surgeons, or whether they place ABPS-certified surgeons, self-described cosmetic surgeons and med spas on the same shortlist. Not a journal, not a certifying board, not a vendor with a disclosed methodology. Anyone claiming AI systematically favors or punishes board certification is guessing, and so is anyone claiming it does not.

What exists is thinner than it sounds. The Plastic Surgery Research Council's 2026 program carries an abstract arguing that large language models are shifting surgeon discovery from search-engine optimization toward recommendation systems (PSRC 2026). It is narrative and conceptual: no sample size, no recommendation rates, no comparison of certified against non-certified surgeons, and it should not be quoted as a measurement. The nearest index ranks brands, not surgeons: 5W Public Relations and Haute MD scored the top 25 medical aesthetics brands off roughly 65 patient-intent prompts (PR Newswire). 5W is a PR and GEO agency selling into this category, so weigh its framing accordingly, and note that no individual surgeon appears in it as a measured entity.

So: documented category confusion exists in exactly the sources an assistant reads, and whether it survives into the answer is unknown for your city. That is a reason to run the check yourself, not to shrug. The answer for your market exists nowhere else.

Four labels an assistant reads as one

The vocabulary is the whole problem. Four titles appear in the same kind of sentence, on the same kind of page, and mean radically different things. A model cannot verify which one it is looking at; it restates what public text says, weighted by how many independent sources say it.

So the move is not to argue with the model about who is a real surgeon. It is to publish the certification as a checkable sentence in crawlable page text: "Certified by the American Board of Plastic Surgery in 2014; plastic surgery residency at [institution]; privileges at [hospital]." A trust seal in the footer does none of that, because it is an image. The GEO study from Princeton, Georgia Tech and IIT Delhi (ACM KDD 2024, 10,000 queries) found citing sources lifted a page's visibility in generated answers by up to 115% for low-ranked content, statistics by 41% and quotations by 28%, while keyword stuffing produced no gain or a loss (arXiv). Specific, sourced claims get repeated; "renowned" does not.

Credential claims inside one clinic's own marketing, who may legally inject, who the medical director is, scope of practice, are a different problem with a different legal surface: see AI visibility for med spas. The problem here is confusion between categories of provider on one shortlist.

• Board-certified plastic surgeon: certified by the American Board of Plastic Surgery, a member board of the American Board of Medical Specialties, after an accredited plastic surgery residency.
• Cosmetic surgeon: a self-description that appears on profiles, practice sites and directory listings. It is not certification by the American Board of Plastic Surgery, whose own 2025 advertising policy governs what its candidates and diplomates may claim (ABPS). Which titles a physician may lawfully advertise is set by state medical boards and varies by state; confirm yours with counsel.
• Aesthetic physician: a description of what someone does, not a certification anyone confers.
• Med spa injector: a label covering whoever a clinic has performing injections. Who may inject, and under whose supervision, is set state by state and the states do not agree: the Illinois IDFPR/IDPH med spa memo updated October 30, 2025 (Illinois) and the Medical Board of California (Medical Board of California) each set out their own rules. Rules vary; confirm your state's with counsel.
• Who may hold which title, and what a physician may advertise, is set by state medical boards and varies by state. Confirm your state's rules with counsel rather than reasoning from another state's.

ABPS restricts "Board Eligible," and California restricts the gallery

The credential language a surgeon most wants is regulated hardest. The American Board of Plastic Surgery's 2025 advertising policy permits "Board Eligible" only after an approved ABPS application, and only for the allowed period defined by the certification process (ABPS, 2025). That is the certifying board's own policy for its candidates and diplomates, not a general law, and it does not tell you what your state medical board separately allows.

The photo gallery is regulated too. California's Business and Professions Code section 651 requires before-and-after images to specify the procedures performed, to be comparable in presentation, and to state that the same results may not occur for all patients; superiority claims must be substantiated with objective scientific evidence (California Legislative Information). That is California law, binding on California practices. Other state medical boards publish their own advertising rules and they differ; whether a specific caption complies in your state is a question for counsel, not a marketing vendor.

There is a quiet upside inside the constraint. A compliant caption names the procedure, describes what was done and carries the results disclaimer, which is almost exactly the crawlable text that makes an otherwise invisible gallery legible to a model. Compliance and visibility are the same keystrokes here. What the rules take away is the superiority claim, "the best facelift in [city]", which California requires you to substantiate scientifically and which an assistant cannot verify anyway.

Tracking your practice name is the wrong unit. Rhinoplasty and gynecomastia are different markets.

A plastic surgery practice does not have one AI visibility score. It has one per procedure, and they diverge. Rhinoplasty, revision rhinoplasty, ethnic rhinoplasty, mommy makeover, facelift, gynecomastia and breast reduction are seven markets with different competitors, different patient language and different sources. A surgeon can own one in the answers and be absent from the other six, and a single check on "best plastic surgeon in [city]" will never show it.

They also pull the assistant toward different sources. Revision patients search by failure rather than proximity and will travel, so the competitor set is national. Gynecomastia is asked about privately, often by someone who has told nobody, which loads the question onto forum text. Breast reduction is half an insurance question, pulling the answer toward payer criteria rather than surgeon directories. "Mommy makeover" is a marketing phrase, not a procedure, so the assistant is matching a phrase, and whoever wrote the clearest page defining it gets quoted.

This is why a practice-level dashboard flatters you: it averages away the one fact worth acting on, which procedure line has zero presence. And the surgical shortlists are the tightest we've measured: in our July 2026 Dallas measurement, assistants named only about 3.9 practices per answer, 43 of the 71 tracked practices (61%) were never named at all, and the 5 most-mentioned practices captured 28% of every recommendation (the Dallas plastic surgery benchmark). A handful of names per question per city. Multiply by seven procedures and the arithmetic stops being comfortable.

The sources AI cites when a patient asks for a rhinoplasty surgeon, ranked by what each one can establish

No one else has published a citation-source study for plastic surgery queries. We have — the one primary measurement in this vertical. In our July 2026 Dallas run — one market, one day, ChatGPT, Gemini and Perplexity — the engines cited 1,153 sources across plastic surgery answers, and 837 of them (73%) were practices' own websites. The largest single third-party platforms were RealSelf at 79, the ASPS site plasticsurgery.org at 31 and Healthgrades at 13, while directories and professional sites — the category that includes the specialty societies and the academic medical centers — drew 161 between them (the Dallas plastic surgery benchmark). Worth noting for a surgeon: in this market the professional and academic sources outweighed every consumer directory, which is not the pattern the listings industry sells. The raw answers behind that count are stored and the method is published, so the run can be checked and repeated line by line. Dallas, July 2026: one city, one day — the pattern, not a national constant.

So the order below is not by citation volume borrowed from other industries. Your own pages come first because the board, the year, the residency and the privileges have to be asserted somewhere before any other source can repeat them — and because that is where the engines went. The rest is assembled from healthcare-wide and local-business-wide data and remains an untested hypothesis for this vertical: the order to check first, not a finding. Verify it against what the engines actually cite in your city.

• 1. Your own website: the only source in this list where the certification can be stated at all — board, year, residency, hospital privileges — and 73% of everything the engines cited in our Dallas run (benchmark). The nearest third-party figure agrees in direction: Gemini takes 52.15% of citations from brands' own sites, ChatGPT nearly the mirror image at 48.73% from third-party listings (Yext, a listings vendor, so a finding that listings decide everything sells listings management — read the emphasis accordingly). Action: one crawlable text page per procedure, each carrying the credential in words. Not a single "Procedures" page behind a dropdown.
• 2. RealSelf and the professional and academic sites (plasticsurgery.org, the specialty societies, the teaching hospitals): the largest third-party sources in our Dallas citations — RealSelf at 79 and ASPS at 31, inside a professional-and-directory category that drew 161 in all (the benchmark) — and the places a credential gets corroborated by somebody who is not you. One city, one day; we make no claim about your metro, and we name no practice. Action: complete the RealSelf and society profiles so the board, the year and the procedure list match your own pages word for word. A fact two independent sources state identically is the closest thing to verification an engine has.
• 3. Reddit and Wikipedia, for ChatGPT: Semrush's AI Visibility Index (126 million prompts, 1,200+ brands, 22 verticals, January to April 2026) found ChatGPT cites 15.4 sources per answer, leaning on Reddit and Wikipedia, against Gemini's 3.3 (Semrush, via PPC Land). Semrush sells AI-visibility tracking. Action: procedure threads name surgeons and you control none of it. Read them; do not manufacture them.
• 4. Zocdoc, for Perplexity: it dominates provider queries there, ahead of Healthgrades and Vitals (rater8 on the Yext data — rater8 sells review-collection software to medical practices, and Yext sells listings management, so both have something to sell here). Action: worth most to a practice doing insured work, breast reduction, gynecomastia, reconstruction; worth less to a pure cash cosmetic practice. Weigh it against your procedure mix.
• 5. Yelp and social profiles, for Google AI Mode: AI Mode cites 11.4 sources per answer and reaches for Facebook, Instagram and Yelp; AI Overviews use 9.2 (Semrush, via PPC Land). Yelp alone drew 512,680 citations across 28 million AI answers in Q4 2025, more than BBB, Angi, Thumbtack, HomeAdvisor and Nextdoor combined (Foundation Inc. with AirOps, a content agency working with an AI-search vendor, so their own tooling is the subject). Action: Instagram is where the 2025 study found the credential problem and engines read it. Put the board in the bio, not just the specialty.
• 6. General healthcare and cosmetic surgery directories (Healthgrades, WebMD, Vitals, and the paid surgery directories): the largest measured share on paper and the smallest in practice. They took 52.6% of healthcare citations in Yext's October 2025 analysis of 6.8 million citations across 1.6 million AI answers (Yext; Yext sells listings management, so a finding that listings decide everything sells listings management), and Perplexity takes 24% of citations on subjective queries from niche industry directories, which is what "who does the most natural rhinoplasty" is. In our Dallas run, directories of every kind drew 8 citations out of 1,153 (benchmark) — one city, one day, so do not retire them on our number alone. Action: claim them, state the ABPS certification and the procedure list on each, and log them as corroboration and hygiene rather than a channel. A profile can confirm a fact or contradict it; it cannot state one your own pages never made.

Seven prompts a patient types, one per procedure. Run them in an afternoon.

Open ChatGPT, Gemini and Perplexity, use a fresh chat per prompt, and paste the questions below with your city in the brackets. Write down every practice named, in order. Then do the step specific to this specialty: next to each name, write what that provider actually is. ABPS-certified plastic surgeon, physician from another specialty using the words "cosmetic surgeon", or a med spa. Certification is verifiable against the certifying board's public records and your state medical board in about a minute per name.

That column is the measurement. If a surgical question returns providers who are not surgeons, you have found in your own market the thing no published study has looked for. If it does not, you have retired a worry with evidence rather than a vendor's assurance. Run it per procedure: a practice named in three of seven and absent from four does not have an "AI visibility problem", it has four dead procedure lines and three live ones, and those need opposite responses.

Two limits. One pass is a snapshot, and answers move between runs, cities and model updates, so a notepad cannot show whether a competitor is pulling ahead over a quarter. And being named is not being described correctly: SOCi's 2026 Local Visibility Index, built on roughly 350,000 locations by a company selling local-marketing software, measured business-data accuracy at about 68% for ChatGPT and Perplexity against 100% for Gemini, which is grounded in Google Maps (SOCi). Getting recommended with the wrong phone number is not a win.

• "Who is the best board-certified rhinoplasty surgeon in [city] for natural results?"
• "I need a revision rhinoplasty specialist with ethnic rhinoplasty experience near [city]. Who should I see?"
• "Which mommy makeover surgeon in [city] has before-and-after photos I can look at?"
• "Gynecomastia specialist near [city]?"
• "Breast reduction in [city]: will insurance cover it, and who performs it?"
• "Who does the most natural-looking facelift in [city]?"
• "Is [your practice name] in [city] a good plastic surgeon?" That one shows what the engines already believe about you.

Questions plastic surgeons ask before changing anything

Nothing in this measurement touches a patient record. It reads public answers to public questions: no EHR, no chart, no photo, no login. Worth saying in a specialty where every software vendor eventually wants a hook into something clinical.

And the honest ceiling: no one can guarantee an assistant will name your practice, for any procedure. Every gap closed raises the probability that the next answer includes you, and probability, not certainty, is the only framing a credential-driven specialty should accept from a marketing vendor, this one included. RecoSignal's free AI Visibility Snapshot runs patient-style prompts across ChatGPT, Gemini and Perplexity on a schedule and reports which practices were named, which sources each answer cited, and where you sit against the surgeons named instead of you. A measurement, not a promise.

• Isn't this just SEO renamed? No. Citations drawn from the organic top 10 fell from 76% to 38% (Ahrefs), an SEO vendor with a commercial interest in AI-search visibility, and SOCi's 2026 index — published by a company that sells local-marketing software and therefore has an interest in the gap it reports — found that in retail only 45% of the top-20 brands in traditional local search reached the top of AI recommendations (Search Engine Land). That gap is the subject of AI visibility versus local SEO.
• Do patients really ask an assistant about surgery? Nobody has measured it for plastic surgery. Healthcare-wide, rater8's June 2026 survey of about 1,000 US adults, run by a vendor selling review-collection software to medical practices, put the share of patients using AI to find a new provider at 47%, up from 31% in late 2025, with heaviest use among 45-to-60-year-olds at 64% (rater8, TechTarget). Carried to a cosmetic practice, that is a hypothesis, not a finding.
• Can I ask patients for reviews that name the procedure? Review text gives an engine more to work with than a star average. But once you publish a patient's story, photo or before-and-after as marketing, it becomes regulated advertising: in California, BPC section 651 governs the images and superiority claims (leginfo). State rules vary, and privacy law is a separate tripwire, documented for a neighboring specialty in the HIPAA review trap. Confirm with counsel before anything goes live.
• The med spa down the street advertises "surgeon-led". What do I do? Nothing that involves characterizing them; we make no claims about any named competitor. What you can do is measure whether an assistant places non-surgical providers on your shortlist for a surgical question. That is a checkable fact about the answer, not an accusation about a business.
• How much works without an agency? Most of the first pass. Rewrite each procedure page as crawlable text, state the ABPS certification in words on every one, claim and correct the healthcare directory profiles, run the seven prompts monthly, keep the notepad. Then decide what is worth paying for, using the ordered playbook of fixes.