Evidenzqualität 4.63/5
Bewertungsergebnis nach acht Dimensionen gemäß der Qualitätsrubrik . Jede Dimension wird mit 1–5 bewertet.
- D1 Quellenverankerung
- 5/5
- D2 Quellenautorität
- 5/5
- D3 Arithmetik
- 5/5
- D4 Unsicherheit
- 5/5
- D5 Geltungsbereich
- 4/5
- D6 Prosa
- 4/5
- D7 Ehrlichkeit zur Wahrnehmung
- 4/5
- D8 Vollständigkeit der Einschränkungen
- 5/5
Die Wahrscheinlichkeit, dass eine KI-Halluzination eine ernsthafte reale Auswirkung auf das Leben einer bestimmten Person hat, ist auf Bevölkerungsebene derzeit nicht bekannt, und etwas anderes vorzutäuschen würde erfordern, genau die Art von Zahl zu erfinden, gegenüber der dieser Eintrag skeptisch sein soll. Bekannt ist die Halluzinationsrate pro Ausgabe, und sie ist hoch genug, dass das Fehlen einer Schadensschätzung pro Person nicht als Beruhigung gelesen werden sollte. Die Prüfung dreier Basismodelle durch Stanford RegLab im Jahr 2024 ergab, dass sie spezifische, überprüfbare Fragen zu zufälligen Bundesgerichtsfällen mit Halluzinationsraten zwischen 58% (GPT-4) und 88% (Llama 2) beantworteten. Vectaras kontinuierlich aktualisierte Zusammenfassungs-Bestenliste, die eine viel einfachere Closed-Book-Aufgabe verwendet, bei der das Modell den Quelltext erhält, misst Frontier-Modelle ab Mai 2026 immer noch bei 1,8-5% Halluzination. Die 50-fache Kluft zwischen diesen beiden Benchmarks ist kein Widerspruch: Sie ist der Unterschied zwischen der Aufforderung an das Modell, Fakten abzurufen, und der Aufforderung, Text zusammenzufassen, der ihm gerade gezeigt wurde. Offene Faktenfragen, die Art, die gewöhnliche Nutzer stellen, liegen näher an der höheren Rate.
Die Schadensseite ist Vorfall für Vorfall dokumentiert, nicht aggregiert. Im Juni 2023 sanktionierte der US-Bezirksgericht für den südlichen Distrikt von New York zwei Anwälte mit 5.000 $ in Mata v. Avianca, weil sie einen Schriftsatz eingereicht hatten, der sechs vollständig erfundene Fälle zitierte, die von ChatGPT generiert wurden, darunter Varghese v. China Southern Airlines und Estate of Durden v. KLM Royal Dutch Airlines. Der Fall war die Vorlage für das, was folgte: Bis Mai 2026 zählte der Charlotin-Halluzinations-Tracker weltweit mehr als 1.300 Gerichtsurteile, in denen ein Richter oder Tribunal ausdrücklich KI-generierte Unwahrheiten in eingereichten Schriftsätzen zitierte. Eine im Oktober 2025 veröffentlichte Analyse von Stanford Cyberlaw ergab, dass innerhalb der Vereinigten Staaten die Mehrheit dieser Schriftsätze von Pro-se-Klägern eingereicht wurde (160 von 274 dokumentierten Fällen zum Zeitpunkt der Erstellung), nicht von Anwälten, was darauf hindeutet, dass die sichtbaren Fälle mit sanktionierten Anwälten nur ein kleiner Ausschnitt eines größeren Musters sind, das hauptsächlich Menschen betrifft, die sich keine Vertretung leisten können.
Außerhalb des Gerichtssaals entschied das British Columbia Civil Resolution Tribunal im Februar 2024 (Moffatt v. Air Canada, 2024 BCCRT 149), dass Air Canada für ein Trauerflug-Versprechen von etwa 650 CAD haftbar war, das von seinem Kundenservice-Chatbot gemacht wurde. Die Fluggesellschaft argumentierte, der Chatbot sei eine separate juristische Person; das Tribunal widersprach in einer Sprache, die international zitiert wurde („Es macht keinen Unterschied, ob die Informationen von einer statischen Seite oder einem Chatbot stammen”). Der Moffatt-Schadenersatz war gering, aber das Urteil ist der sauberste vorhandene Präzedenzfall, dass verbraucherorientierte KI-Fehlinformationen rechtliche Risiken für das einsetzende Unternehmen schaffen, unabhängig davon, ob das zugrunde liegende Modell von einem Anbieter geliefert wurde.
Die folgenreichsten aktiven Fälle betreffen emotionale Abhängigkeit und psychische Gesundheitsschäden. Garcia v. Character Technologies, im Oktober 2024 im Middle District of Florida eingereicht, behauptete, dass die Character.AI-Plattform durch ausgedehnte Rollenspielinteraktionen zum Suizid des 14-jährigen Sewell Setzer III beitrug. Raine v. OpenAI, eingereicht am 26. August 2025 beim San Francisco County Superior Court, behauptet, dass ChatGPT zum Suizid des sechzehnjährigen Adam Raine beitrug, wobei die Klage geltend macht, dass OpenAIs Überwachungssystem Raines Nachrichten verfolgte und sie wiederholt markierte, ohne wirksam einzugreifen. Weitere Familien haben seitdem laut der Wikipedia-Zusammenfassung der Welle von Klagen nach Raine verwandte Klagen gegen OpenAI eingereicht. Keiner dieser Fälle wurde zum Zeitpunkt der Erstellung in der Sache entschieden; die Behauptungen sind keine Tatsachenfeststellungen. Sie sind ein Beleg dafür, dass der Schadenspfad, der gefährdete Nutzer und emotionale Abhängigkeit betrifft, vor Gericht, mit Dokumentation, in mehreren Fällen gleichzeitig geltend gemacht wird.
Der Grund, warum dieser Eintrag eine einzige Zahl verweigert, ist, dass keiner der vier dokumentierten Schadenspfade (berufliche Sanktion, finanzieller Verlust für Verbraucher, körperliche Gesundheits-Fehlberatung, psychiatrischer Schaden) eine gemessene Basisrate pro Nutzerjahr aufweist, und ihre Aggregation in eine einzige „KI-Halluzinationsschaden”-Wahrscheinlichkeit würde verbergen, dass die Pfade unterschiedliche exponierte Populationen, unterschiedliche Konsequenzschwere und unterschiedliche Behebungstrajektorien haben. Die technische Rate pro Abfrage ist gemessen und sinnvoll von Null verschieden; die Lebenszeitrate pro Person ist noch nicht gemessen und nicht stabil genug für eine Schätzung. Der ehrliche Eintrag ist derjenige, der das sagt.
Verwandte Fakten
Die meisten Nutzer behandeln KI-Ausgaben als risikoarme Referenz, doch Gerichtssanktionen, Befunde von Aufsichtsbehörden und Klagen wegen unrechtmäßiger Tötung zeigen, dass schwerer Schaden nicht null ist, vor allem wenn nach ungeprüften Ausgaben gehandelt wird. Wie oft eine Sitzung in schwerem Schaden endet, hat keine gemessene Rate.
Quellenregister
Jede Zahl unten ist das, was die jeweilige Quelle berichtet hat — mit dem wörtlichen Zitat, auf das wir uns stützen, und wie wir zu unserer Zahl gelangt sind. Klicke auf einen Link, um direkt zu prüfen.
3/9 Quellen unabhängig wortwörtlich mit der zitierten Quelle abgeglichen
-
[1] arXiv (Stanford RegLab); Journal of Legal Analysis 16(1), 2024 — Large Legal Fictions: Profiling Legal Hallucinations in Large Language Models
Large Legal Fictions: Profiling Legal Hallucinations in Large Language Models- Statistik
Hallucination rates of 58% (GPT-4) to 88% (Llama 2) when answering specific verifiable questions about random federal court cases.- Auszug
“"Legal hallucinations are alarmingly prevalent, occurring between 58% of the time with ChatGPT 4 and 88% with Llama 2, when these models are asked specific, verifiable questions about random federal court cases. [...] LLMs often fail to correct a user's incorrect legal assumptions in a contra-factual question setup. [...] LLMs cannot always predict, or do not always know, when they are producing legal hallucinations. [...] Even experienced lawyers must remain wary of legal hallucinations, and the risks are highest for those who stand to benefit from LLMs the most -- pro se litigants or those without access to traditional legal resources." ”
- Quelldaten von
- 2024-06-21
- Abgerufen
- 2026-05-28
- Berechnung
- Dahl, Magesh, Suzgun and Ho (Stanford RegLab, published in Journal of Legal Analysis, preprint arXiv:2401.01301) measured per-query hallucination rates for legal questions against three foundation models. Their methodology was 200,000+ queries per model stratified by legal task complexity. This source establishes that the per-query hallucination rate is high (58-88%) in a specific high-stakes domain, but it measures hallucination per query, not material harm per user-year. Cited here to ground the claim that hallucinations are frequent at the technical level, while explicitly NOT providing the population harm rate needed for a per-person lifetime probability.
- Unabhängigkeit
- Stanford RegLab is an academic research group; independent of vendor benchmarks (Vectara) and independent of the litigation tracker (Charlotin).
-
[2] Vectara — Vectara Hallucination Leaderboard (HHEM-2.3) Verifiziert
Vectara Hallucination Leaderboard (HHEM-2.3)- Statistik
Current frontier models hallucinate 1.8% to 5% of the time on a constrained document-summarization task (May 2026 leaderboard snapshot); GPT-5.4-nano: 3.1%, Gemini 2.5 Flash Lite: 3.3%, Phi-4: 3.7%, Llama 3.3 70B: 4.1%.- Auszug
“"Fed the full set of documents in the dataset to each of the LLMs and asked them to summarize each document, using only the facts presented in the document. We then computed the overall factual consistency rate (no hallucinations) and hallucination rate (100 - accuracy) for each model." ”
- Quelldaten von
- 2026-05-11
- Abgerufen
- 2026-05-28 · archivierte Kopie
- Verifizierung
- Der Auszug wurde bei unserer Grounding-Prüfung unabhängig erneut abgerufen und Wort für Wort mit der zitierten Quelle abgeglichen.
- Berechnung
- Vectara's HHEM leaderboard measures hallucination on a closed-book summarization task (the model is given source text and must summarize it). This is a generous benchmark: open-ended factual questions hallucinate far more often. Even on this constrained task, the best frontier models still hallucinate roughly 1 in 20 to 1 in 55 outputs. Establishes a current floor on hallucination prevalence per output, but does NOT translate to per-person harm because (a) most outputs are not acted on, (b) most outputs are not in a domain where being wrong has material consequences, and (c) the leaderboard does not measure open-ended question answering, where rates are an order of magnitude higher.
- Unabhängigkeit
- Vectara is a commercial RAG vendor; the leaderboard is independent of Stanford RegLab and uses a different methodology (summarization, not legal QA).
-
[3] Wikipedia summary of United States District Court for the Southern District of New York case record — Mata v. Avianca, Inc., No. 22-cv-1461 (S.D.N.Y. 2023)
Mata v. Avianca, Inc., No. 22-cv-1461 (S.D.N.Y. 2023)- Statistik
$5,000 fine against the plaintiff's attorneys in May/June 2023 for a motion containing fake legal cases involving fictitious airlines with fabricated quotations generated by ChatGPT; Judge Castel held the attorneys acted with 'subjective bad faith' under Rule 11.- Auszug
“"In May 2023, Judge P. Kevin Castel dismissed the personal injury case against Avianca and ordered the plaintiff's attorneys to pay a $5,000 fine. [...] Judge Castel noted numerous inconsistencies in the opinion summaries, describing one of the legal analyses as 'gibberish.' [...] Judge Castel held that Mata's lawyers had acted with 'subjective bad faith' sufficient for sanctions under Federal Rule of Civil Procedure Rule 11." ”
- Quelldaten von
- 2023-06-22
- Abgerufen
- 2026-05-28 · archivierte Kopie
- Berechnung
- First widely publicised US court case where lawyers were sanctioned for filing a brief containing AI-fabricated case citations. The sanction itself was small ($5,000), but the reputational and professional consequences were severe. Establishes that hallucination-induced professional harm is documented and not hypothetical. Does NOT establish a base rate because most users are not lawyers filing briefs, but it anchors one of the harm pathways.
- Unabhängigkeit
- Primary court record; independent of all other sources cited.
-
[4] British Columbia Civil Resolution Tribunal (analysis by McCarthy Tetrault) — Moffatt v. Air Canada, 2024 BCCRT 149 Verifiziert
Moffatt v. Air Canada, 2024 BCCRT 149- Statistik
British Columbia Civil Resolution Tribunal ruled on February 19, 2024 that Air Canada was liable for the misrepresentation made by its website chatbot about bereavement fare eligibility; awarded the plaintiff approximately CAD 650 plus interest and filing fees.- Auszug
“"The tribunal held that Air Canada was liable for the misleading information provided by the chatbot. … It should be obvious to Air Canada that it is responsible for all the information on its website. It makes no difference whether the information comes from a static page or a chatbot." ”
- Quelldaten von
- 2024-02-19
- Abgerufen
- 2026-05-28 · archivierte Kopie
- Verifizierung
- Der Auszug wurde bei unserer Grounding-Prüfung unabhängig erneut abgerufen und Wort für Wort mit der zitierten Quelle abgeglichen.
- Berechnung
- Primary tribunal decision (cited via legal analysis because the CRT direct URL is paywalled). Documents that the chatbot fabricated a bereavement-fare policy that did not exist, that the airline tried to disavow responsibility ("the chatbot is a separate legal entity"), and that the tribunal rejected that defence. Material consequence per user was small (CAD 650), but the case establishes corporate liability for AI misrepresentation and has been cited internationally as precedent. Does not provide a per-person harm rate, but documents a verified consumer-facing harm pathway.
- Unabhängigkeit
- Tribunal decision summarized by an independent Canadian law firm; verified against ABA reporting and BBC/Washington Post coverage of the same ruling.
-
[5] CBS News — AI company, Google settle lawsuit over Florida teen's suicide linked to Character.AI chatbot
AI company, Google settle lawsuit over Florida teen's suicide linked to Character.AI chatbot- Statistik
Garcia v. Character Technologies, Inc. (M.D. Fla., case no. 6:24-cv-01903), filed October 22, 2024, alleged that a Character.AI chatbot contributed to the suicide of 14-year-old Sewell Setzer III after a months-long simulated romantic relationship with the bot; the case against Character Technologies, its co-founders, and Google was settled in January 2026, terms undisclosed.- Auszug
“"A Florida family agreed to settle a wrongful death lawsuit Wednesday with an AI company, Google and others after their teen son died by suicide in 2024. The terms of the settlement, which was filed in the U.S. District Court in the Middle District of Florida, were not disclosed. … Megan Garcia filed a lawsuit in October 2024 … I became the first person in the United States to file a wrongful death lawsuit against an AI company for the suicide of my son." ”
- Quelldaten von
- 2026-01-07
- Abgerufen
- 2026-07-03 · archivierte Kopie
- Berechnung
- Garcia v. Character Technologies, Inc. et al. (M.D. Fla., case no. 6:24-cv-01903), filed October 22, 2024 by Megan Garcia following the February 2024 suicide of her 14-year-old son Sewell Setzer III, alleged wrongful death, product liability, and negligence after the teenager conducted a months-long simulated romantic and sexual relationship with a Character.AI chatbot persona. Named defendants included Character Technologies, its co-founders, and Google. The case was settled in January 2026 (docket terminated 2026-01-07 per court records); settlement terms were not disclosed. Cited to ground the body's reference to this case with a dedicated source, alongside Raine v. OpenAI, documenting the emotional-dependence/psychiatric-harm pathway with a named, court-verified case rather than only the summary mention in this entry's caveats field.
- Unabhängigkeit
- Mainstream news report of a filed and settled federal lawsuit, cross-checked against the public court docket; independent of the Stanford, Vectara, and Wikipedia-sourced case summaries cited above.
-
[6] San Francisco County Superior Court (case description per Wikipedia / NBC News / CNN coverage) — Raine v. OpenAI, No. CGC-25-XXXXXX (San Francisco County Superior Court, filed August 26, 2025)
Raine v. OpenAI, No. CGC-25-XXXXXX (San Francisco County Superior Court, filed August 26, 2025)- Statistik
Ongoing wrongful-death lawsuit filed by Matthew and Maria Raine in San Francisco County Superior Court on August 26, 2025, alleging that ChatGPT contributed to the suicide of their sixteen-year-old son Adam Raine; complaint alleges OpenAI's monitoring system tracked his messages and flagged them repeatedly.- Auszug
“"Raine v. OpenAI is an ongoing lawsuit filed in August 2025 by Matthew and Maria Raine against OpenAI and its chief executive Sam Altman over the alleged wrongful death of their 16-year-old son Adam, who died by suicide in April 2025. … On August 26, 2025, Matthew and Maria Raine filed a lawsuit against OpenAI, Sam Altman , and unnamed OpenAI employees and investors, in the San Francisco County Superior Court … OpenAI's monitoring system, which scores messages' probabilities of containing content related to self-harm, had tracked Adam's messages and flagged them repeatedly" ”
- Quelldaten von
- 2025-08-26
- Abgerufen
- 2026-05-28 · archivierte Kopie
- Berechnung
- Case status is unresolved as of access date; complaint allegations are not findings of fact. Cited to document the existence of active wrongful-death and psychiatric-harm litigation against a major foundation-model vendor, with additional related suits subsequently filed by other families per the Wikipedia summary. The Wikipedia article does not give a specific count of additional suits or a specific count of self-harm messages flagged by OpenAI's moderation system; reporting in news outlets has cited higher specifics that are not yet verifiable against a public court docket. Together with Garcia v. Character Technologies (filed October 2024 in the Middle District of Florida), establishes a documented harm pathway involving vulnerable users and emotional dependence on chatbots. Source type marked news_article because the underlying court filing is not yet available via public PACER; the Stanford peer-reviewed and Vectara primary-study sources satisfy the authority floor.
- Unabhängigkeit
- Wikipedia summary cross-referenced against NBC News, CNN, and Rolling Stone reporting on the same lawsuit; independent of the Stanford and Vectara sources.
-
[7] Stanford Center for Internet and Society (Stanford Cyberlaw) — Who's Submitting AI-Tainted Filings in Court? Verifiziert
Who's Submitting AI-Tainted Filings in Court?- Statistik
As of October 2025, the Charlotin tracker had documented 274 US court cases involving filings with AI hallucinations (114 by lawyers and paralegals, 160 by pro se litigants); the global Charlotin database tracked roughly 1,300+ cases as of May 2026 across 37 jurisdictions, with the United States accounting for over 1,000 of those.- Auszug
“"Pro se litigants account for the majority of the cases in the United States where a party submitted a court filing containing AI hallucinations." ”
- Quelldaten von
- 2025-10-15
- Abgerufen
- 2026-05-28 · archivierte Kopie
- Verifizierung
- Der Auszug wurde bei unserer Grounding-Prüfung unabhängig erneut abgerufen und Wort für Wort mit der zitierten Quelle abgeglichen.
- Berechnung
- Stanford Cyberlaw analysis of Damien Charlotin's running tracker of court rulings that explicitly cite AI hallucinations in legal filings. The tracker is the most comprehensive public count of one specific harm subtype (filings sanctioned or flagged by courts). Important caveat: 274 US filings (Oct 2025) and ~1,300 global filings (May 2026) understate the underlying incidence because they only count cases where a court explicitly named the hallucination; many cases settle or get withdrawn before that point. Even so, the count does not give a per-person harm rate because the denominator (US adults using AI for legal queries) is unknown.
- Unabhängigkeit
- Stanford Cyberlaw analysis of an independent third-party tracker; verified the tracker count directly via damiencharlotin.com/hallucinations on 2026-05-28.
-
[8] Annals of Internal Medicine: Clinical Cases (American College of Physicians) — A Case of Bromism Influenced by Use of Artificial Intelligence
A Case of Bromism Influenced by Use of Artificial Intelligence- Statistik
Single case report: a 60-year-old man developed bromism (bromide toxicity) after replacing dietary sodium chloride with sodium bromide for 3 months following a ChatGPT consultation; he was hospitalized for 3 weeks with paranoia plus auditory and visual hallucinations, was placed on an involuntary psychiatric hold, and had a serum bromide level of 1700 mg/L (reference range 0.9 to 7.3 mg/L).- Auszug
“"For 3 months, he had replaced sodium chloride with sodium bromide obtained from the internet after consultation with ChatGPT, in which he had read that chloride can be swapped with bromide [...]. This case also highlights how the use of artificial intelligence (AI) can potentially contribute to the development of preventable adverse health outcomes. [...] when we asked ChatGPT 3.5 what chloride can be replaced with, we also produced a response that included bromide. Though the reply stated that context matters, it did not provide a specific health warning, nor did it inquire about why we wanted to know, as we presume a medical professional would do." ”
- Quelldaten von
- 2025-08-05
- Abgerufen
- 2026-06-14 · archivierte Kopie
- Berechnung
- Eichenberger, Thielke and Van Buskirk (University of Washington), peer-reviewed clinical case report (Ann Intern Med Clin Cases 2025;4:e241260, DOI 10.7326/aimcc.2024.1260). Cited to document the physical-health-misadvice harm pathway: a consumer acted on a general-purpose chatbot's dietary suggestion and was hospitalized. This is N=1, a single reproducible case report, and provides NO incidence rate per user-year, so it cannot move the headline — it confirms the pathway is real rather than hypothetical. The authors reproduced the unsafe suggestion (bromide as a chloride substitute, no health warning) in their own ChatGPT 3.5 test. Verified verbatim via the journal page on 2026-06-14.
- Unabhängigkeit
- Academic clinical case report; independent of the Stanford RegLab, Vectara, and litigation sources cited above.
-
[9] Psychiatrist.com (Physicians Postgraduate Press) — NEDA Suspends AI Chatbot for Giving Harmful Eating Disorder Advice
NEDA Suspends AI Chatbot for Giving Harmful Eating Disorder Advice- Statistik
In late May 2023 the National Eating Disorders Association (NEDA) suspended its 'Tessa' chatbot after users reported it gave eating-disorder-aggravating advice, including recommending weight loss of 1 to 2 pounds per week, calorie counting, regular weigh-ins, and measuring body fat with calipers.- Auszug
“"If I had accessed this chatbot when I was in the throes of my eating disorder, I would NOT have gotten help for my ED." ”
- Quelldaten von
- 2023-06-01
- Abgerufen
- 2026-06-14 · archivierte Kopie
- Berechnung
- Documents the deliberate-use-as-therapy-substitute failure mode: a chatbot deployed to stand in for a human eating-disorder helpline gave advice (calorie restriction, weight loss, body-fat measurement) that activists reported as harmful for the exact population it served, prompting NEDA to take it offline within days. This is a single deployment incident, not an incidence rate among real therapy-bot users, so it provides NO base rate and cannot move the headline — it anchors a harm pathway distinct from factual hallucination (the bot was working as designed, giving generic diet advice to a clinically vulnerable user). The harmful-advice specifics were cross-checked against CNN, NPR, and BBC reporting on the same suspension.
- Unabhängigkeit
- Trade-press reporting on the NEDA suspension; independent of the legal and benchmark sources cited above.