The German Medicines Advertising Act and AI Marketing: What Doctors Can and Can't Say

Why AI visibility suddenly matters for doctors

Patients don't just search anymore, they ask. "Which dermatologist in Munich does mole screening with dermoscopy?" or "Is there a cardiologist near me who treats heart rhythm disorders?" Questions like these now go straight into ChatGPT or Perplexity , or surface as the AI summary sitting above Google's results. That answer names specific practices, focus areas, and recommendations. If your practice isn't in it, you don't exist for that patient.

Generative Engine Optimization, GEO for short, is the work of getting into that answer. Unlike classic SEO, it isn't about ranking first in a results list, it's about a language model treating your practice as a trustworthy, verifiable source worth building into its answer. That bar is higher for doctors specifically, because AI providers classify health information as a "Your Money or Your Life" topic and scrutinize it more heavily than most content.

In practice that means AI systems are more skeptical of medical claims by default. They favor sources that are verifiable, factual, and clearly authored. A sweeping promise of a cure doesn't just risk a legal problem, the models tend to filter it out entirely. This is exactly where the German Medicines Advertising Act and AI ranking behavior converge.

What the German Medicines Advertising Act actually prohibits

The HWG regulates advertising for treatments, procedures, and medicines, and three areas matter most for your practice. First, misleading advertising under Section 3: you can't claim an effect that isn't established, and you can't promise a cure with certainty. "We will cure your migraines permanently" is a clear violation. Second, advertising through fear under Section 11, wording built to frighten patients into treatment.

Third, and this one gets missed: the ban on before-and-after imagery for surgical and plastic-surgical procedures, plus restrictions on third-party endorsements and testimonials. Expert opinions, patient quotes, and scientific-sounding language that overstate an effect fall under the same rules. So a dentist can't advertise a staged before-and-after image of a whitening treatment if it edges into surgical-procedure territory.

Here's the distinction worth understanding: the HWG doesn't ban information, it bans promotional, misleading advertising. Factually explaining your services, your qualifications, and your areas of focus is explicitly allowed. And that factual explanation is exactly the raw material AI answers are built from.

The surprising harmony: the HWG and AI want the same thing

Put the two requirements side by side. The HWG wants: factual, verifiable, no exaggeration, clear limits on claimed effect. A language model wants the same thing: verifiable facts, precise language, a clear source, no marketing filler. It's practically one checklist. Where old copy said "the best practice in town," you win with AI systems by saying "dermatology practice specializing in skin cancer screening, using digital dermoscopy since 2018."

The reason comes down to how these models work. They extract entities and facts: specialty, location, procedure, qualification, hours. A line like "pain-free root canals thanks to advanced technology" gives the AI nothing to verify and reads as marketing, so it gets discarded. "Root canal treatment under a dental microscope with mechanical preparation," on the other hand, is a clean, extractable fact, HWG-compliant and AI-friendly at once.

This is the core insight for doctors: you don't have to trade legal safety for visibility. The wording your lawyer signs off on is usually the same wording that earns you a place in the AI answer.

Concrete wording: permitted and prohibited

Look at real examples from practice. Prohibited, and useless to AI: "Finally live pain-free, our therapy is guaranteed to work." That line violates the ban on efficacy guarantees, and AI systems ignore it as an unsubstantiated marketing claim. Permitted and AI-ready: "We treat chronic back pain with multimodal pain therapy, among other approaches. Outcomes vary by patient and are discussed during the initial consultation."

Second example, aesthetics: a before-and-after image of a wrinkle injection captioned "This is how young you'll look afterward" is prohibited. A factual description like "Wrinkle treatment with hyaluronic acid, outpatient, includes consultation on risks and limits" is both permitted and effective. The AI pulls the procedure, the conditions, and your commitment to informed consent straight out of that description, and treats it as a signal of credibility.

Third example, patient testimonials: effusive lines like "Dr. X saved my life, a miracle doctor!" sit in a legal gray zone and read to AI systems as manipulated marketing. Neutral, verifiable Google reviews with real substance, about wait times, communication, and staff friendliness, are the opposite: legally clean and a strong trust signal for the models.

How to make your practice readable for AI

The technical groundwork is structured, machine-readable information. Use schema markup like "MedicalClinic" or "Physician" on your site, with specialty, address, hours, and services listed. That lets the AI understand not just that you exist, but exactly what you offer and to whom. A cardiologist who gives each focus area, echocardiography, long-term ECG monitoring, a heart failure clinic, its own clearly named service page gets cited far more precisely.

Answer real patient questions in plain language on your site. Build an advice section that factually addresses things like "When should I see a cardiologist about heart palpitations?" or "How does a skin cancer screening work?" Question-and-answer content is the format language models favor most, because it can be lifted almost directly into an AI answer. Keep it strictly informative rather than promotional, that protects you under the HWG and improves your findability at the same time.

Keep your listings consistent everywhere: Google Business Profile, jameda, Doctolib, your own site. When name, specialty, and address match across all of them, the models trust your data more. Conflicting details tend to make the AI leave you out entirely rather than risk citing something wrong.

E-E-A-T: why medical authority matters for AI

On health topics, Google and language models both scrutinize source trustworthiness more than usual. The framework they use is called E-E-A-T E-E-A-T: Experience, Expertise, Authoritativeness, Trust. For a doctor this should be an easy win, if you make it visible. On every professional article, name the author with full name, specialist title, and qualifications. An anonymous advice page carries less weight with AI than a text with clear medical authorship.

Link your articles to professional societies, clinical guidelines, or reputable sources such as the AWMF guidelines. When your piece on skin cancer screening references official recommendations, it signals to the AI that you're aligned with scientific consensus. That's legally clean, since you're informing rather than promoting, and it raises your credibility with the models. Those references also lower the odds that the AI treats you as a questionable source.

Press coverage, conference talks, and teaching all add up. When you're quoted in a professional article or write a guest piece, mentions of your name appear elsewhere online. Language models piece together your reputation from many such signals across the web. For doctors, the rule of thumb holds: professional visibility beyond your own website is often worth more than any ad.

The typical mistakes doctors pay dearly for

The most common mistake is trying to impress the AI with superlatives. "Leading practice," "Germany's number one," "revolutionary method," wording like this is exposed under the HWG and is a favorite target for competitors' cease-and-desist letters. It doesn't help with AI systems either, since the models recognize unsubstantiated superlatives as marketing and filter them out. You end up risking a warning letter for zero visibility gain.

The second mistake is handing text creation to a marketing agency with no review for medical advertising law. Copy that reads well promotionally is often a legal minefield under the HWG. Have every piece that mentions treatment outcomes, procedures, or comparisons checked for compliance. The upside: the compliant, defused version is almost always the better GEO text too, because it ends up more factual and fact-dense.

The third mistake is doing nothing. Many doctors assume a good reputation speaks for itself. That's no longer true in an AI-mediated search world. If you don't provide structured, factual information, the AI fills the gap with whatever it finds on review sites, forums, or outdated directories, sources you don't control.

Your roadmap for the coming weeks

Start with an audit. Ask ChatGPT, Gemini and Perplexity specifically about your specialty in your city: "Good orthopedist in Regensburg for sports injuries?" See whether and how you show up, and what's said about you. These answers are your starting point, and they often surface gaps or errors worth fixing right away.

From there, clean up your data foundation: consistent details across every directory, proper schema markup on your site, clearly named service pages, and a factual advice section built around real patient questions. Test every line against two questions: is it HWG-compliant, meaning factual and verifiable? And is it concrete enough for an AI to build an answer from? If both answers are yes, you're on the right track.