AI 生成健康信息的隐形风险
The invisible risks of AI-generated health information
AI 生成健康信息的隐形风险涉及聊天机器人健康查询、症状自查工具准确性和社交媒体误导信息传播。一例溴中毒病例报告称其受 AI 使用影响,另有诉讼指 ChatGPT 导致健康危机。系统综述和调查还关注健康误导信息、疫苗犹豫及 LLM 健康查询使用情况。
Eichenberger, A., Thielke, S. & Van Buskirk, A. A case of bromism influenced by use of artificial intelligence. Ann. Intern. Med. Clin. Cases 4, e241260 https://doi.org/10.7326/aimcc.2024.1260 (2025).
Rosenbluth, T. ChatGPT led to a man’s near-fatal health crisis, lawsuit claims. The New York Times https://www.nytimes.com/2026/07/22/well/openai-chatgpt-health-lawsuit.html (22 July 2026).
Lajka, A., Niu, I., Boyer, M., Surdam, J. & Peters, D. T. The fake influencers selling wellness on your feed. The New York Times https://www.nytimes.com/video/technology/100000011001849/ai-influencers-health-supplements-fake-ads.html (21 July 2026).
Haux, R. Health information systems — past, present, future. Int. J. Med. Inform. 75, 268–281 https://doi.org/10.1016/j.ijmedinf.2005.08.002 (2006).
Wang, X., Shi, J. & Kong, H. Online health information seeking: a review and meta-analysis. Health Commun. 36, 1163–1175 https://doi.org/10.1080/10410236.2020.1748829 (2021).
Borges do Nascimento, I. J. et al. Infodemics and health misinformation: a systematic review of reviews. Bull. World Health Organ. 100, 544–561 https://doi.org/10.2471/BLT.21.287654 (2022).
Krishnan, M. M., Manoharan, G., Selvaraj, F. J. & Shiva, R. K. Harnessing AI for innovative marketing strategies in the healthcare sector. In 2025 IEEE International Conference on Data Science and Information System (ICDSIS) (IEEE, 2025).
Godard, R. & Holtzman, S. Are active and passive social media use related to mental health, wellbeing, and social support outcomes? A meta-analysis of 141 studies. J. Comp. Med. Commun. 29, zmad055 https://doi.org/10.1093/jcmc/zmad055 (2024).
Wallace, W. et al. The diagnostic and triage accuracy of digital and online symptom checker tools: a systematic review. NPJ Digit. Med. 5, 118 https://doi.org/10.1038/s41746-022-00667-w (2022).
Hagger, M. S. & Orbell, S. The common sense model of illness self-regulation: a conceptual review and proposed extended model. Health Psychol. Rev. 15, 494–538 https://doi.org/10.1080/17437199.2021.1878050 (2021).
Walther, J. B., Pingree, S., Hawkins, R. P. & Buller, D. B. Attributes of interactive online health information systems. J. Med. Internet Res. 7, e33 https://doi.org/10.2196/jmir.7.3.e33 (2005).
Swire-Thompson, B. & Lazer, D. Public health and online misinformation: challenges and recommendations. Ann. Rev. Public Health 41, 433–451 https://doi.org/10.1146/annurev-publhealth-040119-094127 (2020).
Obermeyer, Z., Powers, B., Vogeli, C. & Mullainathan, S. Dissecting racial bias in an algorithm used to manage the health of populations. Science 366, 447–453 https://doi.org/10.1126/science.aax2342 (2019).
Pierri, F. et al. Online misinformation is linked to early COVID-19 vaccination hesitancy and refusal. Sci. Rep. 12, 5966 https://doi.org/10.1038/s41598-022-10070-w (2022).
Reid, E. Supercharging search with generative AI. The Keyword ∣ Google Official Blog https://blog.google/products/search/generative-ai-search/ (10 May 2023).
Ayo-Ajibola, O., Davis, R. J., Lin, M. E., Riddell, J. & Kravitz, R. L. Characterizing the adoption and experiences of users of artificial intelligence-generated health information in the United States: cross-sectional questionnaire study. J. Med. Int. Res. 26, e55138 https://doi.org/10.2196/55138 (2024).
Shen, J. H. et al. How people ask Claude for personal guidance. Anthropic https://www.anthropic.com/research/claude-personal-guidance (30 April 2026).
Costa-Gomes, B. et al. Public use of a generalist LLM chatbot for health queries. Nat. Health 1, 689–696 (2026).
Schoenegger, P. et al. Global analysis of country-level factors associated with chatbot usage for health. Nat. Health https://doi.org/10.1038/s44360-026-00174-2 (2026).
OpenAI. Introducing ChatGPT Health. OpenAI https://openai.com/index/introducing-chatgpt-health/ (7 January 2026).
Wachter, R. M. & Brynjolfsson, E. Will generative artificial intelligence deliver on its promise in health care? JAMA 331, 65–69 https://doi.org/10.1001/jama.2023.25054 (2024).
Shah, S. J. et al. Ambient artificial intelligence scribes: physician burnout and perspectives on usability and documentation burden. J. Am. Med. Inform. Assoc. 32, 375–380 https://doi.org/10.1093/jamia/ocae295 (2025).
Anthropic. Advancing Claude in healthcare and the life sciences. Anthropic https://www.anthropic.com/news/healthcare-life-sciences (11 January 2026).
Google Cloud. Google Cloud announces new generative AI advancements for healthcare and life science organizations. Google Cloud Press Corner https://www.googlecloudpresscorner.com/2024-03-12-Google-Cloud-Announces-New-Generative-AI-Advancements-for-Healthcare-and-Life-Science-Organizations (12 March 2024).
Panteli, D. et al. Artificial intelligence in public health: promises, challenges, and an agenda for policy makers and public health institutions. Lancet Public Health 10, e428–e432 (2025).
Mei, K. X., Wolfe, R., Weber, N. & Saveski, M. Grok in the wild: characterizing the roles and uses of large language models on social media. In Proc. International AAAI Conference on Web and Social Media 20, 1609–1633 https://doi.org/10.1609/icwsm.v20i1.42715 (2026).
Renault, T., Mosleh, M. & Rand, D. G. @Grok is this true? LLM-powered fact-checking on social media. Preprint at PsyArXiv https://doi.org/10.31234/osf.io/85quw_v2 (2025).
Costello, T. H. et al. Large language models can effectively convince people to believe conspiracies. Preprint at https://doi.org/10.48550/arXiv.2601.05050 (2026).
DeVerna, M. R., Yan, H. Y., Yang, K. -C. & Menczer, F. Fact-checking information from large language models can decrease headline discernment. Proc. Natl Acad. Sci. USA 121, e2322823121 https://doi.org/10.1073/pnas.2322823121 (2024).
Piccardi, T. et al. Reranking partisan animosity in algorithmic social media feeds alters affective polarization. Science 390, eadu5584 (2025).
Stray, J. et al. The prosocial ranking challenge: reducing polarization on social media without sacrificing engagement. Preprint at https://doi.org/10.48550/arXiv.2603.19626 (2026).
Moynihan, R., Gotzsche, P. C., Heath, I. & Henry, D. Selling sickness: the pharmaceutical industry and disease mongering. BMJ 324, 886–891 (2002).
Saeidnia, H. R., Jahani, S., Ghiasi, N. & Keshavarz, H. Generative AI and health misinformation: production, propagation, and mitigation—a systematic review. BMC Public Health 26, 693 (2026).
Campbel, D. AI deepfakes of real doctors spreading health misinformation on social media. The Guardian https://www.theguardian.com/society/2025/dec/05/ai-deepfakes-of-real-doctors-spreading-health-misinformation-on-social-media (5 December 2025).
Myers, S. L., Callahan, A. & Rosenbluth, T. The doctors are real, but the sales pitches are frauds. The New York Times https://www.nytimes.com/2025/09/05/technology/ai-doctor-scams.html (5 September 2025).
Silverman, C. A real doctor used a pseudonym to post viral health AI slop and sell books. Indicator https://indicator.media/p/a-real-doctor-used-a-pseudonym-to-post-viral-health-ai-slop-and-sell-books (9 March 2026).
Schroeder, D. T. et al. How malicious AI swarms can threaten democracy. Science 391, 354–357 (2026).
DeVerna, M. R. et al. Modeling the amplification of epidemic spread by individuals exposed to misinformation on social media. NPJ Complex. 2, 11 (2025).
Wardle, C., Urbani, S. & Wang, E. Evolving health information-seeking behavior in the context of Google AI overviews, ChatGPT, and Alexa: interview study using the Think-Aloud protocol. J. Med. Internet Res. 27, e79961 https://doi.org/10.2196/79961 (2025).
Fernández-Pichel, M., Pichel, J. C. & Losada, D. E. Evaluating search engines and large language models for answering health questions. NPJ Digit. Med. 8, 153 https://doi.org/10.1038/s41746-025-01546-w (2025).
Hopkins, A. M., Logan, J. M., Kichenadasse, G. & Sorich, M. J. Artificial intelligence chatbots will revolutionize how cancer patients access information: ChatGPT represents a paradigm-shift. JNCI Cancer Spectr. 7, pkad010 https://doi.org/10.1093/jncics/pkad010 (2023).
Montero, A. et al. KFF tracking poll on health information and trust: use of AI for health information and advice. KFF Poll https://www.kff.org/public-opinion/kff-tracking-poll-on-health-information-and-trust-use-of-ai-for-health-information-and-advice/ (25 March 2026).
Astor, M. Doctors couldn’t help them. They rolled the dice with A.I. The New York Times https://www.nytimes.com/2026/04/02/well/live/ai-illness-claude-chatgpt.html (2 April 2026).
Daraz, L. et al. Readability of online health information: a meta-narrative systematic review. Am. J. Med. Qual. 33, 487–492 https://doi.org/10.1177/1062860617751639 (2018).
Gotlieb, R. et al. Accuracy in patient understanding of common medical phrases. JAMA Netw. Open 5, e2242972 https://doi.org/10.1001/jamanetworkopen.2022.42972 (2022).
Romoff, M., Brunette, M., Peterson, M. K., Hashmi, S. Z. & Kim, M. S. The role of large language models in improving the readability of orthopaedic spine patient educational material. J. Orthop. Surg. Res. 20, 531 https://doi.org/10.1186/s13018-025-05955-1 (2025).
Hoek, A. E. et al. Patient discharge instructions in the emergency department and their effects on comprehension and recall of discharge instructions: a systematic review and meta-analysis. Ann. Emerg. Med. 75, 435–444 https://doi.org/10.1016/j.annemergmed.2019.06.008 (2020).
Huang, L. et al. A survey on hallucination in large language models: principles, taxonomy, challenges, and open questions. ACM Trans. Inf. Syst. 43, 1–55 https://doi.org/10.1145/3703155 (2025).
Ramaswamy, A. et al. ChatGPT Health performance in a structured test of triage recommendations. Nat. Med. 32, 1671–1675 (2026).
Cacioli, M., Arya, A., Liao, A. & Zhu, K. Do knowledge cutoffs drive clinical accuracy? Quantifying temporal decay in large language models. In Socially Responsible and Trustworthy Foundation Models at NeurIPS 2025 (NeurIPS, 2025).
DeVerna, M. R., Yang, K. -C., Yan, H. Y. & Menczer, F. Large language models require curated context for reliable political fact-checking—even with reasoning and web search. In Findings of the Association for Computational Linguistics 29338–29360 (Association for Computational Linguistics, 2026)
Omar, M. et al. Mapping the susceptibility of large language models to medical misinformation across clinical notes and social media: a cross-sectional benchmarking analysis. Lancet Digit. Health 8, 100949 https://doi.org/10.1016/j.landig.2025.100949 (2026).
Hu, D. et al. Auditing Google’s AI Overviews and featured snippets: a case study on baby care and pregnancy. In Proceedings of the International AAAI Conference on Web and Social Media Vol. 20, 1044–1062 https://doi.org/10.1609/icwsm.v20i1.42681 (2026).
Yang, K. -C. & Menczer, F. Accuracy and political bias of news source credibility ratings by large language models. In Proc. 17th ACM Web Science Conference, Websci ’25, 127–137 https://doi.org/10.1145/3717867.3717903 (Association for Computing Machinery, 2025).
Alber, D. A. et al. Medical large language models are vulnerable to data-poisoning attacks. Nat. Medicine 31, 618–626 https://doi.org/10.1038/s41591-024-03445-1 (2025).
Abtahi, F., Seoane, F., Pau, I. & Vega-Barbas, M. Data poisoning vulnerabilities across health care artificial intelligence architectures: analytical security framework and defense strategies. J. Med. Int. Res. 28, e87969 https://doi.org/10.2196/87969 (2026).
Stokel-Walker, C. Scientists invented a fake disease. AI told people it was real. Nature https://doi.org/10.1038/d41586-026-01100-y (2026).
Marks, M. & Haupt, C. E. AI chatbots, health privacy, and challenges to HIPAA compliance. JAMA 330, 309–310 https://doi.org/10.1001/jama.2023.9458 (2023).
Fernández-Pichel, M., Losada, D. E. & Pichel, J. C. Large language models for binary health-related question answering: a zero- and few-shot evaluation. In Computational Science – ICCS 2024 (eds Franco, L. et al.) 325–339 https://doi.org/10.1007/978-3-031-63772-8_29 (Springer Nature, 2024).
Udry, J. & Barber, S. J. The illusory truth effect: a review of how repetition increases belief in misinformation. Curr. Opin. Psychol. 56, 101736 https://doi.org/10.1016/j.copsyc.2023.101736 (2024).
Loomba, S., de Figueiredo, A., Piatek, S. J., de Graaf, K. & Larson, H. J. Measuring the impact of COVID-19 vaccine misinformation on vaccination intent in the UK and USA. Nat. Hum. Behav. 5, 337–348 (2021).
Bollenbacher, J., Menczer, F. & Bryden, J. Effects of antivaccine tweets on COVID-19 vaccinations, cases, and deaths. EPJ Data Sci. 15, 12 (2026).
Ferrara, E. GenAI against humanity: nefarious applications of generative artificial intelligence and large language models. J. Comput. Soc. Sci. 7, 549–569 (2024).
Costello, T. H., Pennycook, G. & Rand, D. G. Durably reducing conspiracy beliefs through dialogues with AI. Science 385, eadq1814 (2024).
Xu, H. G. et al. Personalized AI dialogues and parents’ HPV vaccine intent: a randomized trial. Preprint at PsyArXiv https://doi.org/10.31234/osf.io/gv52j_v1 (2025).
Jin, Y. et al. Better to ask in English: cross-lingual evaluation of large language models for healthcare queries. In Proc. ACM Web Conference 2024, WWW ’24, 2627–2638 https://doi.org/10.1145/3589334.3645643 (Association for Computing Machinery, 2024).
Freelon, D., Monzer, C., Jeon, G., Moy, C. & Williams, N. The post-API age of social media data access: past, present, and future. Ann. Am. Acad. Polit. Soc. Sci. 715, 16–37 (2025).
Ortutay, B. Meta kills off misinformation tracking tool Crowdtangle despite pleas from researchers, journalists. The Associated Press https://apnews.com/article/meta-crowdtangle-research-misinformation-shutdown-facebook-977ece074b99adddb4887bf719f2112a (15 August 2024).
Mimizuka, K., Brown, M. A., Yang, K. -C. & Lukito, J. Post-post-API age: studying digital platforms in scant data access times. In Proc. ACM on Human-Computer Interaction https://doi.org/10.1145/3816985 (Association for Computing Machinery, 2026).
Pan, J. et al. COMPASS-GH is a consensus roadmap for defining standards for safe, accurate and equitable AI in general health queries. Nat. Health 1, 162–163 (2026).
Essén, A. et al. Patient access to electronic health records: differences across ten countries. Health Policy Technol. 7, 44–56 https://doi.org/10.1016/j.hlpt.2017.11.003 (2018).
Beninger, P. Pharmacovigilance: an overview. Clin. Ther. 40, 1991–2004 https://doi.org/10.1016/j.clinthera.2018.07.012 (2018).
Dangy-Caye, A. et al. Harmonizing health: a global analysis of pharmaceutical regulatory activities by international regulatory organizations. Front. Med. 12, 1636269 https://doi.org/10.3389/fmed.2025.1636269 (2025).
Angus, D. C. et al. AI, health, and health care today and tomorrow: the JAMA summit report on artificial intelligence. JAMA 334, 1650–1664 (2025).
Goanta, C. et al. The great data standoff: researchers vs. platforms under the digital services act. In Proc. 20th International AAAI Conference on Web and Social Media 20, 874–888 (Association for the Advancement of Artificial Intelligence, 2026).
US Food and Drug Administration. Prescription-drug advertisements. 21 C.F.R. § 202.1. FDA & Department of Health and Human Services https://www.ecfr.gov/current/title-21/chapter-I/subchapter-C/part-202/section-202 .1 (accessed 16 August 2026).
Leonardo Alves, T., Lexchin, J. & Mintzes, B. Medicines information and the regulation of the promotion of pharmaceuticals. Sci. Eng. Ethics 25, 1167–1192 https://doi.org/10.1007/s11948-018-0041-5 (2019).
European Parliament and Council of the European Union. Regulation (EU) 2024/1689 (Artificial Intelligence Act). OJ L, 2024/1689, 12.7.2024. CELEX: 32024R1689 4. http://data.europa.eu/eli/reg/2024/1689/oj (accessed 16 August 2026).
Transparency Coalition. German court holds Google liable for AI hallucination: read the full decision here. Transparency Coalition https://www.transparencycoalition.ai/news/german-court-holds-google-liable-for-ai-hallucination-read-the-full-decision-here (15 June 2026).
Ong, J. C. L. et al. Ethical and regulatory challenges of large language models in medicine. Lancet Digit. Health 6, e428–e432 (2024).
Meskó, B. & Topol, E. J. The imperative for regulatory oversight of large language models (or generative AI) in healthcare. NPJ Digit. Med. 6, 120 (2023).
来源:Nature 健康护理主题 · nature.com