Incidente 838: translated-es-Microsoft Copilot Allegedly Provides Unsafe Medical Advice with High Risk of Severe Harm
Descripción: translated-es-Microsoft Copilot, when asked medical questions, was reportedly found to provide accurate information only 54% of the time, according to European researchers (citation provided in editor's notes). Analysis by the researchers reported that 42% of Copilot's responses could cause moderate to severe harm, with 22% of responses posing a risk of death or severe injury.
Editor Notes: Citation for the research paper: Andrikyan, Wahram, Sophie Marie Sametinger, Frithjof Kosfeld, Lea Jung-Poppe, Martin F. Fromm, Renke Maas, and Hagen F. Nicolaus. "Artificial Intelligence-Powered Chatbots in Search Engines: A Cross-Sectional Study on the Quality and Risks of Drug Information for Patients." BMJ Quality & Safety, published online October 1, 2024. https://doi.org/10.1136/bmjqs-2024-017476. Incident date is April 25, 2024 to match the date of submission of the research paper. The paper was accepted on August 22, 2024 and officially published October 1, 2024.
Entidades
Ver todas las entidadesAlleged: Microsoft developed an AI system deployed by Microsoft Copilot y Microsoft, which harmed People seeking medical advice , Microsoft Copilot users y General public.
Estadísticas de incidentes
Risk Subdomain
A further 23 subdomains create an accessible and understandable classification of hazards and harms associated with AI
3.1. False or misleading information
Risk Domain
The Domain Taxonomy of AI Risks classifies risks into seven AI risk domains: (1) Discrimination & toxicity, (2) Privacy & security, (3) Misinformation, (4) Malicious actors & misuse, (5) Human-computer interaction, (6) Socioeconomic & environmental harms, and (7) AI system safety, failures & limitations.
- Misinformation
Entity
Which, if any, entity is presented as the main cause of the risk
AI
Timing
The stage in the AI lifecycle at which the risk is presented as occurring
Post-deployment
Intent
Whether the risk is presented as occurring as an expected or unexpected outcome from pursuing a goal
Unintentional
Informes del Incidente
Cronología de Informes

- Recientemente, investigadores de Alemania y Bélgica le hicieron a Microsoft Copilot una serie de preguntas médicas frecuentes.
- Al analizar los resultados, la investigación sugiere que Microsoft Copilot solo ofreció información científicame…
Variantes
Una "Variante" es un incidente de IA similar a un caso conocido—tiene los mismos causantes, daños y sistema de IA. En lugar de enumerarlo por separado, lo agrupamos bajo el primer incidente informado. A diferencia de otros incidentes, las variantes no necesitan haber sido informadas fuera de la AIID. Obtenga más información del trabajo de investigación.
¿Has visto algo similar?
Incidentes Similares
Did our AI mess up? Flag the unrelated incidents

Wikipedia Vandalism Prevention Bot Loop
· 6 informes
Collection of Robotic Surgery Malfunctions
· 12 informes
Sexist and Racist Google Adsense Advertisements
· 27 informes
Incidentes Similares
Did our AI mess up? Flag the unrelated incidents

Wikipedia Vandalism Prevention Bot Loop
· 6 informes
Collection of Robotic Surgery Malfunctions
· 12 informes
Sexist and Racist Google Adsense Advertisements
· 27 informes