The New York Times published a fabricated quote. An AI tool generated a summary of Conservative leader Pierre Poilievre's views on Canadian politics and rendered it as a direct quotation. The reporter used it without verification.
The correction is specific: Poilievre did not call politicians who changed allegiances turncoats in his April speech. The Times replaced the hallucinated quote with an accurate one only after the error was discovered post-publication. The editors' note names the failure directly: the reporter should have checked what the AI returned.
Read the original editors' note because the mechanism matters more than the outcome. This is not a case of bias or bad faith. It is a workflow failure where AI output was treated as a primary source. That is the story worth understanding before it happens at your publication.
[READ ORIGINAL →]