An editor at the newspaper reportedly called it “truly astonishing” that the product was launched at all
According to Semafor’s Thursday report based on internal communications at the US newspaper, The Washington Post’s new AI-powered personalized podcasts provided subscribers with fabricated quotes and factual inaccuracies.
Launched earlier this week, the feature provides mobile app users with AI-created podcasts that automatically condense and read aloud chosen news stories, using the newspaper’s written content.
Just 48 hours after the product debuted, WaPo staff started identifying numerous issues, such as made-up quotes, misattributed statements, and inaccurate factual information.
“It is truly astonishing that this was allowed to go forward at all,” one WaPo editor allegedly wrote in an internal message. At the time of Semafor’s report, the WaPo had not publicly recognized the issue.
These reported mistakes occur during increased examination of US media trustworthiness. In late last month, the White House introduced a media bias tracker on its official site, designed to publicly catalog news articles and organizations the administration deems biased or erroneous. The WaPo appears prominently on the tracker along with outlets like CNN, CBS, and Politico.
The Washington Post is considered one of America’s top national newspapers, together with The New York Times and The Wall Street Journal. Amazon founder Jeff Bezos has owned it since 2013. During his ownership, the Post has broadened its digital activities and made substantial investments in technology.
The problems with the WaPo’s AI-produced podcasts also coincide with other major media organizations adopting comparable technologies. Firms such as Yahoo and Business Insider have recently unveiled or expanded AI-powered tools meant to summarize articles, representing a wider industry trend toward employing artificial intelligence to reduce expenses, accelerate production, and tailor content for audiences.
This incident underscores wider worries about artificial intelligence usage in journalism, where automated systems have consistently generated mistakes, known as hallucinations, and deceptive material. Media groups and specialists have cautioned that lacking robust editorial protections, AI-created content threatens to erode precision, responsibility, and public confidence.