AI-Generated Quotes Lead to Suspension of Prominent Journalist
Peter Vandermeersch, a former editor-in-chief of both De Standaard and NRC, has been suspended by Mediahuis, the publisher of the Irish Independent and Sunday Independent, after admitting to using artificial intelligence (AI) to generate fabricated quotes in his work. The incident highlights the growing risks associated with the use of AI writing tools in journalism and other fields.
Background and Career
Vandermeersch recently held the position of Fellow “Journalism and Society” at Mediahuis, a role he took on five months ago following a period as Ireland chief executive from 2022 to 2025. Prior to that, he served as publisher of the Irish organization, known then as Independent News & Media, from 2019 to 2022. He was also editor-in-chief of NRC for nine years in the 2010s.
The Investigation and Admission
The suspension followed an investigation by NRC, a Mediahuis publication, which revealed that Vandermeersch had published “dozens of quotes” that proved to be false. A journalist at NRC initially raised concerns after being unable to verify certain quotes used by Vandermeersch. Analysis of his blog posts – both Dutch and English language versions – showed that 15 out of 53 posts contained quotes that could not be found in the original sources cited, such as news articles and scientific studies. Seven individuals cited in the posts confirmed they had not made the attributed statements, either in the publications mentioned or elsewhere.
Vandermeersch publicly acknowledged his mistake in a Substack article titled “I am admitting my mistake.” He stated he had “place words in people’s mouths.”
Mediahuis Response
Mediahuis has removed eight of Vandermeersch’s articles from the independent.ie website and app. The company’s investigation was prompted by the findings of the NRC investigation.
Broader Implications
This incident is part of a growing trend of errors arising from the use of AI tools like ChatGPT as writing assistants. Previous incidents have occurred in media, science, and the legal profession. Vandermeersch himself had previously expressed concerns about the reliability of AI chatbots, warning in October 2025 that they often provide incorrect answers to current events questions, describing a “hollow core” beneath their smooth delivery. He had also advocated for the eventual elimination of human editors once AI reached a sufficient level of capability.
Looking Ahead
The case of Peter Vandermeersch serves as a cautionary tale about the potential pitfalls of relying on AI-generated content, even for experienced journalists. It underscores the critical importance of rigorous fact-checking and editorial oversight in the age of artificial intelligence.