GEO in Healthcare: How Generative Engine Optimization Can Fight Medical Misinformation

In short: Social media spreads health information based on what gets engagement. AI search tools build their answers from sources they consider reliable. That difference gives healthcare experts something new: if credible content is structured well, AI systems can find it, trust it and cite it. Generative Engine Optimization (GEO) is how you do that. It won’t fix health misinformation alone, but it’s one of the few strategies that work with the new information system rather than against it.

Why is health misinformation so hard to beat on social media?

The problem is how social media works, not only what people post. A study published in Science tracked about 126,000 stories on Twitter. False news spread farther, faster, deeper and more broadly than true news in every category studied. People, not bots, were mainly responsible, partly because false news felt more novel ¹.

The platforms’ ranking systems make this worse. A preregistered audit of Twitter (now X)’s engagement-based algorithm found that it amplified emotionally charged, hostile content more than a simple chronological feed did, and users said they didn’t prefer the posts it picked ².

Health topics follow the same pattern. A University of Sydney-led study looked at 982 Instagram and TikTok posts about five medical tests that can lead to overdiagnosis, including full-body MRI. The accounts had a combined reach of about 194 million followers. 87% of the posts mentioned benefits, fewer than 15% mentioned harms and only about 6% cited scientific evidence ³. In 2023, Kim Kardashian also urged her 360 million followers to undergo “life-saving” full-body MRI despite no evidence of benefit for people at average risk of disease, and the clear danger of overdiagnosis and overuse ³.

A WHO-supported review of systematic reviews found that health misinformation can lead to misread evidence, harm to mental health, misallocated health resources, vaccine hesitancy and delayed care ⁴.

On social media, the most-shared content wins. In healthcare, that is rarely the most accurate content.

How is AI changing the way people look for health information?

Search habits are changing fast. In the first four months of 2026, 68% of US Google searches ended without a click to any website, up from about 60% in 2024. AI Overviews are a likely contributor, though the study didn’t measure how much ⁵.

Health is part of this shift. A BCG survey of 13,353 internet-connected adults in 15 countries found that nearly 60% already use generative AI for their personal health, with especially high use in emerging markets ⁶. For many patients, a chatbot now gives the first answer, before any clinician does.

Patients have reasons to like it. In a JAMA Internal Medicine study, licensed healthcare professionals compared chatbot and physician answers to real patient questions without knowing which was which. They preferred the chatbot’s answers 78.6% of the time and rated them as higher quality and more empathetic ⁷.

Is AI simply the next misinformation machine?

Not necessarily, but it isn’t flawless either. A randomized study in Nature Medicine with 1,298 participants found that LLMs identified the correct condition in 94.9% of scenarios when tested alone. Members of the public using the same models identified it in fewer than 34.5% of cases, which was no better than the control group ⁸. The Oxford researchers saw failures on both sides: users left out key details, and the models sometimes gave misleading answers ⁹.

AI is still risky. What has changed is the kind of risk. Social media’s main weakness is amplifying whatever gets attention. AI’s main weaknesses are errors and the way people interact with it, and better source material and better design can reduce both.

Why does AI favour trusted sources?

This is the central point. Google says its AI Overviews and AI Mode are built on its core Search ranking and quality systems. They use retrieval-augmented generation, which finds relevant pages in the search index and grounds the answer in them, with links back to the sources ¹⁰.

Those quality systems set a higher bar for “Your Money or Your Life” topics such as health. Google’s quality rater guidelines judge content on Experience, Expertise, Authoritativeness and Trust (E-E-A-T), and Trust matters most ¹¹. Google’s guidance for creators asks site owners to make it clear who wrote the content and why readers should trust it ¹².

In principle, then, the system rewards credibility, not popularity. A small website with expert authors and clear sourcing can be cited ahead of a high-traffic site with weaker content. Each AI platform selects sources differently, and none of them is perfect. Still, the logic is very different from a social feed.

What is GEO, and does it work?

The term Generative Engine Optimization comes from a Princeton-led paper presented at KDD 2024. On the authors’ test benchmark, adding citations, statistics and quotations made content up to 40% more visible in AI-generated answers. Keyword stuffing didn’t help ¹³.

Google itself treats GEO as SEO and says there are no extra requirements for appearing in its AI features ¹⁰. In practice, GEO is SEO done properly for the AI era: a disciplined way of making credible content easy for machines to find, understand and cite.

What should medical societies and academic institutions do now?

1. Answer real questions. Write for the question patients actually ask (“why do I wake at 3 a.m.?”), not the textbook topic. Use questions as headings and give a direct answer first.
2. Be specific about evidence. Name the study design, the journal and the sample size rather than “researchers found.” Precise, sourced statements are easier for AI to verify and cite ¹³.
3. Show who is behind the content. Name the author and the medical reviewer, and include their credentials. Add structured data (schema.org) so machines can link people, organizations, conditions and treatments ¹⁴.
4. Be transparent. Publish an editorial policy, review dates, disclosures and any use of AI in writing the content. Trust is built from evidence you show ¹¹.
5. Stay balanced. Cover benefits and harms. That is exactly what the misleading posts in the Sydney study left out ³.

The bottom line

Ignoring AI or simply fearing it won’t protect the public. In healthcare, the real choice isn’t between “Dr Google” and “Dr AI.” It’s about whether trusted human expertise stays visible in a world where AI increasingly mediates health information. GEO is one way to make sure it does.

References

1. Vosoughi S, Roy D, Aral S. The spread of true and false news online. Science. 2018;359(6380):1146–1151. https://doi.org/10.1126/science.aap9559
2. Milli S, et al. Engagement, user satisfaction, and the amplification of divisive content on social media. PNAS Nexus. 2025;4(3):pgaf062. https://doi.org/10.1093/pnasnexus/pgaf062
3. Nickel B, et al. Social media posts about medical tests with potential for overdiagnosis. JAMA Netw Open. 2025;8(2):e2461940. https://doi.org/10.1001/jamanetworkopen.2024.61940
4. Borges do Nascimento IJ, et al. Infodemics and health misinformation: a systematic review of reviews. Bull World Health Organ. 2022;100(9):544–561. https://doi.org/10.2471/BLT.21.287654
5. Goodwin D. Google zero-click searches hit 68% in early 2026: study. Search Engine Land. June 9, 2026. https://searchengineland.com/google-zero-click-searches-2026-study-479717
6. Keneally B, et al. Consumers are ready for AI-enabled health care. Health systems need to be, too. BCG. April 21, 2026. https://www.bcg.com/publications/2026/consumers-are-ready-for-ai-health-care-are-systems
7. Ayers JW, et al. Comparing physician and artificial intelligence chatbot responses to patient questions posted to a public social media forum. JAMA Intern Med. 2023;183(6):589–596. https://doi.org/10.1001/jamainternmed.2023.1838
8. Bean AM, Payne RE, Parsons G, et al. Reliability of LLMs as medical assistants for the general public: a randomized preregistered study. Nat Med. 2026. https://doi.org/10.1038/s41591-025-04074-y
9. University of Oxford. New study warns of risks in AI chatbots giving medical advice. February 10, 2026. https://www.ox.ac.uk/news/2026-02-10-new-study-warns-risks-ai-chatbots-giving-medical-advice
10. Google Search Central. Optimizing your website for generative AI features on Google Search. https://developers.google.com/search/docs/fundamentals/ai-optimization-guide
11. Google. Search Quality Rater Guidelines: an overview. https://services.google.com/fh/files/misc/hsw-sqrg.pdf
12. Google Search Central. Creating helpful, reliable, people-first content. https://developers.google.com/search/docs/fundamentals/creating-helpful-content
13. Aggarwal P, et al. GEO: Generative Engine Optimization. Proceedings of KDD 2024:5–16. https://doi.org/10.1145/3637528.3671900
14. Schema.org. https://schema.org/

Olivier Gryson, PharmD, MSc
25 years of experience in digital marketing in the pharmaceutical industry
Special focus on AI Search in Pharma Marketing