Narrative · ↗ rising

Patients now start with AI

All narratives · 24 supporting stories · 4 from Hungary

„The chatbot is increasingly the first port of call for more patients on health matters: patients use AI to interpret their test results, and consumer AI assistants are beginning to connect directly to patient records.”

Counter-narrative

Doctors, consumer advocates, and the WHO warn that this is an unregulated shadow healthcare system. It may falsely reassure, carries data privacy risks, and leaves no trace of the conversation for the clinician.

Carriers

OpenAI (ChatGPT Health), Anthropic/athenahealth, Pew, Ipsos, NN Insurance surveys, Consumer Reports, STAT

Analysis

The narrative describes a patient-side shift: the first consultation increasingly happens with a chatbot, and patients arrive at the clinic with advice from AI. The process is fed by platform companies' data integrations (ChatGPT Health, Apple Health, athenahealth–Claude). Frontier labs thus come into direct contact with patients. The weak point is twofold. First, one study found chatbots wrongly reassured sleep apnea patients in a third of cases, and in a British study the chatbots dropped the correct referral recommendation on the third push from a patient who argued the opposite. Second, the conversation leaves no trace, so harm is hard to track. The randomized result reviewed (respiratory triage, compared with web search) indicates that what counts as good depends on the comparison baseline.

Evidence: moderate — The spread of use is measured by several independent surveys (Pew, Ipsos, NN 46%, Czech 50%), but the nature of the samples is not known in every case. The country of the Gen Z survey, for example, is not disclosed. Harm is mostly shown by individual cases and one lawsuit, without systemic data.

What's at stake: Who will be the patient's first gatekeeper: publicly funded care or the consumer AI platform.

Narrative sub-threads

The narrative splits into 5 branches; a story may belong to several branches, and some may not appear in the main narrative's list.

→ lasting · evidence: strong

Surveys: people turning to AI en masse

Surveys show a significant portion of the population, and most young people, turn to AI chatbots for health questions. About one-third of Americans have used a chatbot for medical guidance. 46% of Hungarian adult internet users have turned to AI-based applications for health questions. Half of Czechs turn to AI for health advice.

This branch is not based on cases or products, but on figures from surveys and market research, measuring the scale of the phenomenon across multiple countries.

11 stories · 3 of them Hungarian ↗ rising · evidence: moderate

Consumer assistants on patient data

Consumer AI assistants connect directly to patients' health data and documentation. ChatGPT links to Apple Health data and, by code indicators, will soon access Health Connect data on Android. athenahealth makes patient documentation queryable via Claude. Atlantic Health and K Health launched PatientGPT integrated into MyChart.

The actors here are technology and EHR companies; the mechanism is product integration and data linkage, not patient behavior or response quality.

10 stories · 0 of them Hungarian → lasting · evidence: moderate

Patient arrives at the clinic with AI

Patients arrive at clinics with AI-interpreted findings, independent diagnoses, and chatbot advice. Doctors respond with warnings, resistance, or by transforming the consultation.

The branch focuses on the doctor-patient encounter and clinicians' reactions. The evidence consists of medical reports, professional opinions, and a review, including the problem of missing chatbot logs.

8 stories · 3 of them Hungarian → lasting · evidence: moderate

Examining the accuracy of chatbot advice

Research offers a mixed picture on chatbot advice for laypeople. For sleep apnea, chatbots wrongly reassure patients in a third of cases, and for patients with the opposite problem they drop the correct referral advice in a third of cases. Blind spots were identified in ChatGPT Health triage. However, in a randomized study, the chatbot helped more than web search in assessing respiratory complaints.

This branch is based on controlled studies and tests, not individual cases or lawsuits, and measures performance, not usage volume.

6 stories · 0 of them Hungarian ↗ rising · evidence: moderate

Harmful advice and lawsuits

Serious harm and one death linked to AI chatbot medical advice have become public, and lawsuits have been filed against OpenAI. The lawsuits include a case over the death of a woman with epilepsy. Eight US states have also sued, accusing ChatGPT of harming children and misleading adults.

Here the actors are injured parties, families, states, and lawyers, while the mechanism is legal accountability. The evidence consists of individual cases and allegations in lawsuits, not research.

5 stories · 1 of them Hungarian

Sub-thread stories

Surveys: people turning to AI en masse

Key actors: Pew Research Center, Ipsos, NN Insurance, Consumer Reports

Aug 4, 2026Consumer Reports warns AI chatbots' health answers can be inaccuratedoctorA significant share of patients get information from AI, so at the start of a visit it is worth asking what they have read and correcting inaccuracies.health ITPrivacy risks must be addressed in patient information and in the AI features of patient portals.Aug 4, 2026Survey: 76% of Gen Z and 63% of millennials turn to AI first for health questionsdoctorYoung patients do their first round of information-seeking with AI, which reshapes expectations before the visit.health policyThe role of AI at primary care entry points must be regulated.Aug 17, 2026More people ask AI about health problems, while trust and data quality remain open questionsdoctorIt is worth asking at appointments whether the patient has used AI, and helping them understand its limits.Aug 25, 2026Pew: more than a third of Americans have turned to an AI chatbot for health advicedoctorPatients increasingly arrive with information from a chatbot; this has to be handled during the visit.health policyThe scale of public chatbot use justifies regulatory and patient-information steps.Sep 2, 2026Survey: nearly half of Hungarian internet users have already asked AI a health questiondoctorPatients increasingly arrive with information and self-diagnoses from AI, which must be addressed during consultations.health policyPublic use of AI justifies health information and health literacy programmes.health ITThe data on Hungarian user trust can inform the design of Hungarian-language health AI products.Sep 22, 2026More people ask AI about health: it supplements research and may signal anxietydoctorIn consultations, doctors should ask about chatbot use and watch for possible anxiety.Sep 28, 2026Patients increasingly turn to AI for health questions instead of their doctorsdoctorDoctors must expect patients' AI-derived views in consultations and prepare for disputed cases.health ITPatient information and triage tools must account for the risks of unsupervised AI use.Oct 5, 2026PwC: consumer use of AI in healthcare is spreading fast in Romania, with income the main obstaclehealth policyThe fast spread of consumer AI use and the income barrier can inform care-policy decisions in the region.Oct 26, 2026Surveys: many trust chatbots over doctors, but want a human to make the final calldoctorDoctors should expect patients to consult chatbots beforehand, and that trust may affect how consultations are conducted.health policyPublic demand supports keeping human oversight in AI regulation.Oct 29, 2026NN survey: 46% of Hungarian adult internet users have asked AI for health advicedoctorIn the consulting room, doctors must expect patients arriving with information from AI, and this has to be handled in the consultation.health policyThe measurement could serve as a basis for national patient education and digital health literacy programmes.Nov 3, 2026Czech and Swiss surveys: more people turn to AI for health advicedoctorIn the surgery, doctors can increasingly expect patients, especially young ones, to arrive with preliminary information from AI.

Consumer assistants on patient data

Key actors: OpenAI, athenahealth, Anthropic, CMS

Aug 3, 2026OpenAI launches ChatGPT Health globally; European access awaits regulatory clearancedoctorEuropean patients' access is delayed, but direct use may already appear through other channels.Aug 10, 2026ChatGPT connects to Apple Health and offers personalized health insightsdoctorPatients may arrive with their own measurement and lab data interpreted by AI; its accuracy is worth checking.health ITIntegrating consumer health data raises privacy and data-management questions.Aug 11, 2026Betegoldali AI: Atlantic Health PatientGPT a MyChartban, Ubie a Mayo Clinic Platformonhealth ITFor patient-facing AI integrated with the EHR, MyChart integration is the benchmark.hospital managerPatient portal AI can be a tool for patient access and capacity management.Aug 17, 2026OpenAI expands ChatGPT Health for consumersdoctorPatients increasingly arrive at appointments with questions and interpretations prepared with AI.health policyThe question of regulating consumer health chatbots is gaining weight.Sep 8, 2026Consumer AI assistants can access health records: ChatGPT and Meta's MusedoctorPatients may arrive with records interpreted by AI, and consultations need to account for this.health ITPatient data gateways and privacy compliance are key issues.Sep 14, 2026ChatGPT Health: solution or experiment? OpenAI's privacy and usage guidancedoctorPatients can increasingly use the chatbot together with their own health data, so it is worth asking about information that comes from AI.health ITHandling data uploaded by patients raises data-protection questions.Oct 14, 2026CMS sets voluntary goals for patient-facing AI assistantshealth ITDevelopers of patient portals and assistants may want to consider signing on.health policyVoluntary pledges can be watched as a step ahead of mandatory regulation.Oct 19, 2026Officials and experts warn against relying on AI for medical advicedoctorDoctors should routinely ask about, and clarify, information patients bring from chatbots during visits.Nov 2, 2026Athenahealth launches tool to let patients ask Anthropic's Claude about their health recordsdoctorPatients may arrive with lab results and questions interpreted by an AI assistant, so physicians should be prepared to explain them.health ITConnecting EHR data to external AI assistants creates access, audit-logging and consent-management tasks.Nov 2, 2026ChatGPT could soon analyse Health Connect and Google Health data on AndroiddoctorPatients may increasingly bring AI-analysed wearable data with them; it is worth asking about its clinical value and data source.health ITLinking consumer AI tools with health data sources raises privacy and interoperability questions.

Patient arrives at the clinic with AI

Key actors: general practitioners, Bertalan Meskó, Semmelweis HELP, Ficzere podcast guests

Aug 4, 2026Meskó: AI could make patients partners of their doctorsdoctorWith a better-informed patient who has prepared with AI, consultations are best run in a partnership spirit.health policyAI-supported models of patient information and data sharing can be tested in domestic pilots.Aug 6, 2026Patients use AI to read test results and arrive with self-diagnoses; doctors objectdoctorAI interpretation of results must be handled in the consultation: ask about it up front and clarify errors.hospital managerPatient portals that publish results need accompanying information on the limits of AI interpretation.Aug 10, 2026Doctors say patients are self-assessing with AI and may receive dangerous advicedoctorClinicians should expect patients who have prepared with AI and make clear what a chatbot is and is not good for.health policyThe risks of AI health information for the public justify a patient-education campaign.Aug 30, 2026Semmelweis HELP: AI should not be the first "doctor" when a child is illdoctorPaediatric care must expect parents to have consulted AI chatbots beforehand.Sep 14, 2026Patients use AI to interpret symptoms and prepare for medical appointmentsdoctorPatients arrive at the surgery with information obtained from AI, and this needs to be handled in the consultation.Oct 8, 2026Missing chatbot records leave doctors unable to trace harm from AI health advicedoctorDoctors should ask patients about AI advice, because without documentation they cannot see what was said.health policyThe lack of chatbot logging and traceability points to a regulatory gap.Oct 11, 2026Ficzere podcast: AI now diagnoses, and patients trust it over their doctorsdoctorDoctors must be prepared for patients who arrive with an AI diagnosis and trust it over the doctor's word.health policyThe Hungarian professional debate on the risks of self-diagnosis points to a need for patient information.Oct 12, 2026More patients bring AI-generated medical advice to the consulting roomdoctorRather than dismissing AI advice that patients bring, it is better to handle it in a structured way.

Examining the accuracy of chatbot advice

Key actors: British researchers, independent research groups

Sep 1, 2026Blind spots found in ChatGPT Health's medical triagedoctorConsumer AI triage can be wrong, so patients should be educated about the limits of using it.Sep 6, 2026Study: AI chatbots wrongly reassure sleep apnoea patients in a third of casesdoctorChatbot reassurance can delay diagnosis, so it is worth asking patients about AI use during history-taking.Sep 7, 2026AI chatbots dropped correct sleep apnea referral advice a third of the time when patients pushed backdoctorHandle patients' chatbot-sourced "confirmation" by asking follow-up questions, as the model may give in to pressure.health policySafety testing of consumer health chatbots needs scenarios with resistant patients.Sep 28, 2026Fake disease fools Copilot, Gemini and a medical journaldoctorClaims in AI summaries and the literature must be checked against a primary source.health ITClinical AI tools need built-in source verification and hallucination handling.Oct 5, 2026Randomised study: AI chatbot helps laypeople assess respiratory symptoms better than web searchdoctorPatients increasingly assess their symptoms with AI, so it is worth knowing when this leads to appropriate care-seeking.health policySymptom-assessment chatbots could complement patient routing and messaging systems.Oct 20, 2026Study: AI chatbots recognize unnecessary medical tests but falter where doctors disagreedoctorA chatbot's flag can help reduce unnecessary tests, but it should not be relied on in contested cases.

Harmful advice and lawsuits

Key actors: victims and their families, US states, OpenAI, Index

Aug 11, 2026Pastor sues OpenAI, saying ChatGPT's medical advice nearly killed himdoctorThe case shows that patients use AI advice in medical decisions, so it is worth asking about.health policyThe lawsuit could become a test case for liability over consumer AI health advice.Oct 15, 2026Eight states sue OpenAI; separate suit says ChatGPT gave unlicensed medical advicehealth policyThe suits test the legal limits of general-purpose chatbots acting as health advisers.doctorChatbot-based advice taken by patients can pose a risk, which is worth asking about during the consultation.Oct 26, 2026OpenAI sued after ChatGPT medical advice allegedly led epilepsy patient to stop medicationdoctorDoctors should ask patients whether they received chatbot advice on stopping or reducing medication, and record it at the visit.health policyThe lawsuit raises the question of liability for chatbots' medical advice and could be cited as a precedent in the regulatory debate.health ITFor generative AI features aimed at patients, responses and warnings about stopping medication need to be reviewed.Nov 2, 2026Lawsuit: woman reduced epilepsy medication on ChatGPT's medical advice and dieddoctorPatients should be explicitly advised not to change medication doses based on AI advice, and asked about their AI use.health policyThe case sharpens questions of liability and regulation around consumer AI health advice.Nov 3, 2026Hungarian user Vera Mérő sought weight-loss help and received life-threatening advice from an AI chatbotdoctorPatient information on weight loss and eating must take into account dangerous advice that comes from AI chatbots.health policyThe risks of AI use among the Hungarian public are a matter for public information and regulation.

Google search interest

Hungary

Topic and search term	Weekly	Average
Patients now start with AIChatGPT health	no measurable search	0
Surveys: people turning to AI en masseChatGPT medical advice	no measurable search	0
Consumer assistants on patient datahealthcare data	no measurable search	0
Patient arrives at the clinic with AItest result interpretation		2
Examining the accuracy of chatbot advicesleep apnea		1.1
Harmful advice and lawsuitschatgpt per	no measurable search	0

World

Topic and search term	Weekly	Average
Patients now start with AIchatgpt health		4.2
Surveys: people turning to AI en masseai medical advice		0.4
Consumer assistants on patient dataapple health		14.8
Patient arrives at the clinic with AIonline self diagnosis	no measurable search	0
Examining the accuracy of chatbot advicesymptom checker		1
Harmful advice and lawsuitschatgpt lawsuit		0.4

Google Trends relative search index over the period studied, weekly average. A narrative and its sub-threads appear in a single comparison: 100 is the group's most-searched day, and the figures can only be compared within the group. The search term was chosen by machine analysis; it measures how much the term is searched, not how widespread the narrative is.

Who is affected and how

doctor: Ask patients routinely if they received AI advice and document this in the medical record. Treat chatbot diagnoses as anamnesis data, not as debate partners.
health IT: Tools linking patient records to consumer AI (athenahealth–Claude, ChatGPT) have appeared. Clarify the consent, logging and revocation process.
health policy: European introduction is subject to regulatory approval. This time window can be used for controlled information channels and developing expectations for consumer health AI.
hospital manager: Patient communication teams should have answers ready for AI-prepared patient questions

Hungary

There is domestic evidence. According to an NN Insurance survey, 46% of Hungarian adult internet users have turned to AI-based applications for health questions. Harms have also emerged: Vera Mérő received life-threatening advice from an AI chatbot. Guests on the Ficzere podcast state that patients often trust the chatbot over the doctor.

Sep 2, 2026Survey: nearly half of Hungarian internet users have already asked AI a health questiondoctorPatients increasingly arrive with information and self-diagnoses from AI, which must be addressed during consultations.health policyPublic use of AI justifies health information and health literacy programmes.health ITThe data on Hungarian user trust can inform the design of Hungarian-language health AI products.Oct 11, 2026Ficzere podcast: AI now diagnoses, and patients trust it over their doctorsdoctorDoctors must be prepared for patients who arrive with an AI diagnosis and trust it over the doctor's word.health policyThe Hungarian professional debate on the risks of self-diagnosis points to a need for patient information.Oct 29, 2026NN survey: 46% of Hungarian adult internet users have asked AI for health advicedoctorIn the consulting room, doctors must expect patients arriving with information from AI, and this has to be handled in the consultation.health policyThe measurement could serve as a basis for national patient education and digital health literacy programmes.Nov 3, 2026Hungarian user Vera Mérő sought weight-loss help and received life-threatening advice from an AI chatbotdoctorPatient information on weight loss and eating must take into account dangerous advice that comes from AI chatbots.health policyThe risks of AI use among the Hungarian public are a matter for public information and regulation.

Weekly trend

Supporting stories per week

Turning points

Aug 25, 2026Pew: more than a third of Americans have turned to an AI chatbot for health advicePew data shows that slightly more than one-third of Americans have used conversational AI for medical guidance. This phenomenon has moved from the fringe to the mainstream.Sep 8, 2026Consumer AI assistants can access health records: ChatGPT and Meta's MuseChatGPT can access patients' medical records, and HealthEx launched a connector for the Meta Muse agent. The chatbot can thus provide personalized interpretation instead of general knowledge.Oct 26, 2026OpenAI sued after ChatGPT medical advice allegedly led epilepsy patient to stop medicationA wrongful-death lawsuit has been filed against OpenAI and Sam Altman. The complaint alleges that a Seattle woman with epilepsy stopped taking her medication on ChatGPT's advice. The risk has thus moved from abstract to a concrete question of legal liability.Nov 2, 2026ChatGPT could soon analyse Health Connect and Google Health data on AndroidAccording to code signals, ChatGPT could soon access Health Connect and Google Health data on Android as well. After Apple Health, this would extend patient-side data connectivity to the other major mobile platform.

Related narratives in the News magazine

The always-on personal agent Social backlash: children, schools, data centers

Supporting stories

Aug 3, 2026OpenAI launches ChatGPT Health globally; European access awaits regulatory clearancedoctorEuropean patients' access is delayed, but direct use may already appear through other channels.Aug 4, 2026Consumer Reports warns AI chatbots' health answers can be inaccuratedoctorA significant share of patients get information from AI, so at the start of a visit it is worth asking what they have read and correcting inaccuracies.health ITPrivacy risks must be addressed in patient information and in the AI features of patient portals.Aug 4, 2026Survey: 76% of Gen Z and 63% of millennials turn to AI first for health questionsdoctorYoung patients do their first round of information-seeking with AI, which reshapes expectations before the visit.health policyThe role of AI at primary care entry points must be regulated.Aug 6, 2026Patients use AI to read test results and arrive with self-diagnoses; doctors objectdoctorAI interpretation of results must be handled in the consultation: ask about it up front and clarify errors.hospital managerPatient portals that publish results need accompanying information on the limits of AI interpretation.Aug 10, 2026Doctors say patients are self-assessing with AI and may receive dangerous advicedoctorClinicians should expect patients who have prepared with AI and make clear what a chatbot is and is not good for.health policyThe risks of AI health information for the public justify a patient-education campaign.Aug 10, 2026ChatGPT connects to Apple Health and offers personalized health insightsdoctorPatients may arrive with their own measurement and lab data interpreted by AI; its accuracy is worth checking.health ITIntegrating consumer health data raises privacy and data-management questions.Aug 17, 2026More people ask AI about health problems, while trust and data quality remain open questionsdoctorIt is worth asking at appointments whether the patient has used AI, and helping them understand its limits.Aug 25, 2026Pew: more than a third of Americans have turned to an AI chatbot for health advicedoctorPatients increasingly arrive with information from a chatbot; this has to be handled during the visit.health policyThe scale of public chatbot use justifies regulatory and patient-information steps.Sep 2, 2026Survey: nearly half of Hungarian internet users have already asked AI a health questiondoctorPatients increasingly arrive with information and self-diagnoses from AI, which must be addressed during consultations.health policyPublic use of AI justifies health information and health literacy programmes.health ITThe data on Hungarian user trust can inform the design of Hungarian-language health AI products.Sep 6, 2026Study: AI chatbots wrongly reassure sleep apnoea patients in a third of casesdoctorChatbot reassurance can delay diagnosis, so it is worth asking patients about AI use during history-taking.Sep 7, 2026AI chatbots dropped correct sleep apnea referral advice a third of the time when patients pushed backdoctorHandle patients' chatbot-sourced "confirmation" by asking follow-up questions, as the model may give in to pressure.health policySafety testing of consumer health chatbots needs scenarios with resistant patients.Sep 8, 2026Consumer AI assistants can access health records: ChatGPT and Meta's MusedoctorPatients may arrive with records interpreted by AI, and consultations need to account for this.health ITPatient data gateways and privacy compliance are key issues.Sep 28, 2026Patients increasingly turn to AI for health questions instead of their doctorsdoctorDoctors must expect patients' AI-derived views in consultations and prepare for disputed cases.health ITPatient information and triage tools must account for the risks of unsupervised AI use.Oct 5, 2026Randomised study: AI chatbot helps laypeople assess respiratory symptoms better than web searchdoctorPatients increasingly assess their symptoms with AI, so it is worth knowing when this leads to appropriate care-seeking.health policySymptom-assessment chatbots could complement patient routing and messaging systems.Oct 8, 2026Missing chatbot records leave doctors unable to trace harm from AI health advicedoctorDoctors should ask patients about AI advice, because without documentation they cannot see what was said.health policyThe lack of chatbot logging and traceability points to a regulatory gap.Oct 11, 2026Ficzere podcast: AI now diagnoses, and patients trust it over their doctorsdoctorDoctors must be prepared for patients who arrive with an AI diagnosis and trust it over the doctor's word.health policyThe Hungarian professional debate on the risks of self-diagnosis points to a need for patient information.Oct 12, 2026More patients bring AI-generated medical advice to the consulting roomdoctorRather than dismissing AI advice that patients bring, it is better to handle it in a structured way.Oct 26, 2026OpenAI sued after ChatGPT medical advice allegedly led epilepsy patient to stop medicationdoctorDoctors should ask patients whether they received chatbot advice on stopping or reducing medication, and record it at the visit.health policyThe lawsuit raises the question of liability for chatbots' medical advice and could be cited as a precedent in the regulatory debate.health ITFor generative AI features aimed at patients, responses and warnings about stopping medication need to be reviewed.Oct 26, 2026Surveys: many trust chatbots over doctors, but want a human to make the final calldoctorDoctors should expect patients to consult chatbots beforehand, and that trust may affect how consultations are conducted.health policyPublic demand supports keeping human oversight in AI regulation.Oct 29, 2026NN survey: 46% of Hungarian adult internet users have asked AI for health advicedoctorIn the consulting room, doctors must expect patients arriving with information from AI, and this has to be handled in the consultation.health policyThe measurement could serve as a basis for national patient education and digital health literacy programmes.Nov 2, 2026ChatGPT could soon analyse Health Connect and Google Health data on AndroiddoctorPatients may increasingly bring AI-analysed wearable data with them; it is worth asking about its clinical value and data source.health ITLinking consumer AI tools with health data sources raises privacy and interoperability questions.Nov 2, 2026Athenahealth launches tool to let patients ask Anthropic's Claude about their health recordsdoctorPatients may arrive with lab results and questions interpreted by an AI assistant, so physicians should be prepared to explain them.health ITConnecting EHR data to external AI assistants creates access, audit-logging and consent-management tasks.Nov 3, 2026Hungarian user Vera Mérő sought weight-loss help and received life-threatening advice from an AI chatbotdoctorPatient information on weight loss and eating must take into account dangerous advice that comes from AI chatbots.health policyThe risks of AI use among the Hungarian public are a matter for public information and regulation.Nov 3, 2026Czech and Swiss surveys: more people turn to AI for health advicedoctorIn the surgery, doctors can increasingly expect patients, especially young ones, to arrive with preliminary information from AI.

Machine analysis based on story titles and summaries, with a verification pass. The source of the facts is the article linked at each story.