Incident library · Media · United States

    CNET: 41 of 77 AI articles corrected

    A technology news site quietly published finance explainers written with an AI tool. A review found errors in 41 of the 77 articles.
    Talking Biz NewsJan 2023

    CNET found errors in more than half of its AI stories

    41 of 77 AI articles corrected

    Stated by the organisation

    AI-written articles were published without review catching factual errors, and readers were not clearly told AI was involved.

    Stated by the organisation

    The reported contributing factor

    AI-written articles were published without review catching factual errors, and readers were not clearly told AI was involved.

    Incidents usually have several contributing factors. This is the one public reporting names; it is not presented as the sole cause.

    Which StackUp assessment checks for this gap

    Find out if your organisation has this gap

    Sources

    Published Jan 2023. Sources last checked Nov 2026. Spotted an error or an update? Tell us and we will correct it.

    • Ars TechnicaApr 2025

      Company apologizes after AI support agent invents policy that causes user uproar

      Public apology and cancellations

      Stated by the organisation

      An AI agent answered customers directly without a clear label or human check, and invented a company policy.

    • Silicon RepublicJan 2025

      Apple calls a halt to AI notification summaries amid backlash

      AI feature suspended after complaints

      Stated by the organisation

      AI summaries were shown to users as if they came from the publisher, without a clear warning that they could be wrong.

    • The MarkupMar 2024

      NYC's AI Chatbot Tells Businesses to Break the Law

      Chatbot later shut down

      Reported contributing factor

      A public chatbot gave legal answers that were wrong, and stayed online with a disclaimer after the errors were reported.

    Find out which of these gaps you have

    The free Readiness Check takes ten questions and shows the incidents tied to your weakest answers.