European Journalist Suspended For Using False AI-Generated Quotes

Dutch media house Mediahuis has suspended its former editor-in-chief, Peter Vandermeersch, for using AI-generated quotes that wrongly “put words into people’s mouths”.

Mediahuis, the publisher of Dutch newspaper De Telegraaf, and the Irish Independent suspended the former head of the Irish operations for ‘falling into the trap of hallucinations’ – the term for AI-generated errors – when using the technology.

In a message on his Substack site, Vandermeersch admitted he had summarised reports using AI tools like ChatGPT, Perplexity, and Google’s NotebookLM, without checking whether the quotes from those summaries were accurate before publishing them in his Substack newsletter.

An investigation by one of Mediahuis’ own titles, NRC (where Vandermeersch had been editor-in-chief in the 2010s), uncovered the errors. NRC alleged that Vandermeersch had published dozens of false quotes and that seven quoted individuals in his posts said they had not made the statements attributed to them.

“I wrongly put words into people’s mouths when I should have presented them as paraphrases. In some cases, it reflected my interpretation of their words. That was not just careless, it was wrong. I am admitting my mistake.”

He continued, “It is particularly painful that I made precisely the mistake I have repeatedly warned colleagues about. These language models are so good that they produce irresistible quotes you are tempted to use as an author. Of course, I should have verified them. The necessary ‘human oversight’ which I consistently advocate fell short.”

He also partially blamed his fascination with AI and his need to experiment extensively with its capabilities.

Mediahuis Group Chief Executive Gert Ysebaert said the medium applies strict rules for AI use, particularly when diligence, human oversight, and transparency are essential.

“The fact that these principles were not followed runs counter to the standards we uphold and to our commitment to readers that we stand for reliable journalism. We are discussing this with Peter Vandermeesch and have decided to suspend him from his role as a fellow temporarily.”

Despite the current outcome, Vandermeersch remains insistent that AI can help improve journalism.

“Journalism is human work. I remain convinced that AI can be a powerful tool, helping journalism become better, dig deeper, and be more precise. But not by using it in the way I did in the early months of this blog.”