An editor at the newspaper reportedly said it was “truly astonishing” the product was launched at all
The Washington Post’s new AI-based personalized podcasts presented subscribers with invented quotes and factual errors, Semafor reported on Thursday citing internal correspondence at the US newspaper.
Rolled out earlier this week, the feature offers mobile app users AI-generated podcasts that automatically summarize and narrate selected news stories, drawing on the newspaper’s written articles.
Within 48 hours of the product launch, WaPo employees began flagging multiple problems, including fabricated quotations, wrongly attributed statements, and incorrect factual details.
”It is truly astonishing that this was allowed to go forward at all,” one WaPo editor reportedly said in an internal message. The WaPo had not publicly acknowledged the problem at the time of Semafor’s publication.
The reported errors come amid heightened scrutiny of US media credibility. Late last month, the White House launched a media bias tracker on its official website, aimed at publicly listing news articles and outlets the administration considers biased or inaccurate. The WaPo features prominently on the site alongside outlets such as CNN, CBS, and Politico.
The Washington Post is regarded as one of the leading US national newspapers, alongside The New York Times and The Wall Street Journal. It has been owned by Amazon founder Jeff Bezos since 2013. Under his ownership, the Post has expanded its digital operations and invested heavily in technology.
The issues surrounding the WaPo’s AI-generated podcasts also come as other major media outlets move to introduce similar technologies. Companies including Yahoo and Business Insider have recently announced or expanded AI-driven tools designed to summarize articles, part of a broader push across the media industry to use artificial intelligence to cut costs, speed up production, and personalize content for readers.
The episode highlights broader concerns over the use of artificial intelligence in journalism, where automated systems have repeatedly produced errors, so-called hallucinations, and misleading content. Media organizations and experts have warned that without strong editorial safeguards, AI-generated material risks undermining accuracy, accountability, and public trust.
