The study identified multiple systemic issues across four leading AI tools. Based on data from 18 countries and 14 languages, 45% of all AI answers had at least one significant issue, and 31% of responses showed serious sourcing problems — missing, misleading, or incorrect attributions.
The PDF report has a lot more detail, including individual examples and analysis of the issues encountered.
Comments
A summary of the study by NPR, who themselves participated in the study, is [here](https://www.npr.org/sections/npr-extra/2025/10/21/g-s1-94424...).
The PDF report has a lot more detail, including individual examples and analysis of the issues encountered.
Also this associated toolkit: https://www.ebu.ch/files/live/sites/ebu/files/Publications/M...