The publisher of the Dutch newspaper De Telegraaf and the Irish Independent has suspended one of its senior journalists after he admitted using AI to “wrongly put words into people’s mouths”.
Peter Vandermeersch, the former head of the Irish operations at Mediahuis, said he “fell into the trap of hallucinations” – the term for AI-generated errors – when using the technology.
Vandermeersch, a fellow of “journalism and society” at the European publishing group, has been suspended from his role.
The experienced journalist said he had summarised reports using AI tools such as ChatGPT, Perplexity and Google’s NotebookLM, and not checked whether the quotes from those summaries were accurate. He subsequently published them in his Substack newsletter.
The errors were highlighted by an investigation by one of Mediahuis’s own titles, NRC, where Vandermeersch had been editor-in-chief in the 2010s. NRC alleged Vandermeersch had published “dozens” of quotes that were false 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,” Vandermeersch wrote in a Substack post headlined “I am admitting my mistake”.
Vandermeersch added: “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.”
Vandermeersch’s Press and Democracy blog writes regularly about “the vital connection between a free press and a healthy democracy”.
In a statement, the Mediahuis group’s chief executive, Gert Ysebaert, said: “At Mediahuis we apply strict rules for the use of AI, where 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 Vandermeersch and have decided to temporarily suspend him from his role as fellow.’’
Mediahuis has also removed a number of articles written by Vandermeersch from the Irish Independent website.
AI tools such as ChatGPT are used by hundreds of millions of people for an array of tasks ranging from the relatively straightforward, such as suggesting recipes, to carrying out complex academic research. However, they are prone to making errors.
Vandermeersch said he made a second mistake by failing to correct false quotes immediately, instead leaving that work to a title he had overseen for nearly a decade. He said he was enthusiastic about the possibilities of AI and had wanted to experiment with them extensively.
“Journalism is human work,” he wrote. “I remain convinced that AI can be a powerful tool – one that can help journalism become better, dig deeper, and be more precise. But not by using AI in the way I did in the early months of this blog.”
Vandermeersch declined to comment.
