AIXDAabout 3 hours ago
AI Summaries Fabricate Report Details
1 min read
ChatGPT, Claude, and Gemini all fabricated details when summarizing a 47-page Reuters Institute report. Gemini made the most errors, misclassifying survey data as measured results. ChatGPT cited sources not in the report, while Claude contradicted the report's measured Chartbeat data.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
