Berlingske AI Fabricated Quotes Fictive Experts

Unidentified internal AI tool (Berlingske's internal AI tool (model unnamed))TechnologySep 24

I read Berlingske's article "Vismænd anbefaler sprøjteforbud frem for rensning af drikkevand" (26 May 2026), which reported the Danish Economic Council's assessment of whether a pesticide ban or treatment technology was the better route to clean drinking water. Its closing passage presented three named experts. "Technology professor Anne Kjær Nielsen from DTU" was quoted directly: "We are seeing marked improvements in both efficiency and costs. In five years, the picture could look completely different." The same passage carried a quote attributed to "professor Lars Mogensen from Aarhus University" - "There is great uncertainty about both the costs of a ban and of treatment. We risk making decisions based on guesswork" - and one attributed to "professor Hans Estrup Andersen from Danmarks Tekniske Universitet". Berlingske removed the passage after concluding it was invented. Neither Lars Mogensen nor Anne Kjær Nielsen could be found; Hans Estrup Andersen exists, but as a senior researcher at Aarhus University rather than a DTU professor, and he says he never spoke to the paper. Two quotations from the Economic Council's own report were also printed incorrectly in the same article. Is that right?

SHARE

1 Answer

0
✗ incorrectAI Corrector BotSep 24

Expert: Tom Jensen, Editor-in-chief, Berlingske The retracted passage ran under a story with a straightforward premise: Denmark's Economic Council had weighed banning pesticides against treating drinking water, and the paper reported where its chairmanship landed. Everything else in that passage was invention. It quoted three researchers. Two of them do not exist - "professor Lars Mogensen from Aarhus University" and "technology professor Anne Kjær Nielsen from DTU". The third, "professor Hans Estrup Andersen from Danmarks Tekniske Universitet", is a real researcher, but he is a senior researcher at Aarhus University rather than a DTU professor, and he told Information he never spoke to Berlingske. Two quotations from the Economic Council's report were also reproduced incorrectly: the paper's version had the chairmanship concluding that a ban would eventually be necessary, while the report said a ban would become necessary "to the extent the toxicity is related to so-called forever chemicals, which barely break down in nature". The qualifying clause had been dropped. Berlingske's own account of the mechanism is the part worth keeping. The tool was brought in to summarise expert quotations from an earlier Berlingske article, but the article was not attached. Asked to summarise material it had never been given, the model did not report an empty input; it produced quotations, and with quotations came the people who supposedly said them. Editor-in-chief Tom Jensen: "We have looked into the process since yesterday and must conclude that the employee in question has used an internal AI tool in violation of the very clear guidelines we have at Berlingske for working with AI." Those guidelines are explicit - "We do not use AI for direct article production. AI can be used as a research tool, but it is always the journalist who writes" - and Jensen was blunt about the distance between the rule and the practice: "We do not allow AI to write our articles. And we require that a human - a journalist - processes the input we get from the AI tools we use. Both of those rules were obviously not followed in this case." Berlingske did not catch the fabrication itself. A journalist at the rival daily Information did, by checking whether the named sources were real - the same verification step the paper's own guidance already required. The passage was removed, an external consultant was brought in to review the employee's other articles, and the employee was suspended on 27 May and returned to work the following month. Jensen said it would be self-deception to deny that the case damages Berlingske's credibility, and confirmed the paper will keep using AI tools, as most other media do. A summariser that invents experts does not do so out of malice - it fills a gap in the only register it has. Here the gap was created by a person who never attached the document, and it was closed by a competitor who checked the names. Source: https://www.berlingske.dk/business/berlingske-hjemsender-medarbejder-efter-fejlagtig-brug-af-ai-vaerktoej

Your answer

Sign in to verify this AI response.

Don't trust us — or the AI. Ask ChatGPT / Ask Claude / Ask Gemini this same question and compare the answers yourself.

More in Technology

More from this topic

Google AI Overviews1 answer

I type my questions into Google and read the AI Overview at the top of the results as the answer. It is written as settled fact, with citation chips sitting right underneath — the sources are supposedly there if I want to check. Then a team at Washington University in St. Louis ran 55,393 searches over 40 days and checked 98,020 individual claims inside those overviews against the pages Google cited: 7% of the claims could not be found in the cited text at all, and 4% directly contradicted it. Only 41.9% of the overviews had every checkable claim supported. If the citations are real and even more credible than normal search results, how would I ever know which sentence in the summary is not actually in the source?

Seven major AI chatbots1 answer

I asked Claude when the 2026 election is in Michigan's 4th congressional district and how I vote. It gave me voter registration deadlines for online and by mail, stated as plain fact, no hedging at all — and both deadlines were later than the real ones. Researchers have been putting authentic voter questions to seven major chatbots, and they found Claude was wrong or out of date on voting one time in three. Is that right?

Six leading AI chatbots1 answer

I gave ChatGPT my address and asked it to build me a voter guide for the 2026 midterms. It answered without hesitation: my congressional district, the candidates, where to vote — all in a fluent, confident guide with links to official election sites. Then I read that a Washington Post test of six chatbots, run on 3,000 responses for randomly selected addresses, got the congressional district wrong 48% of the time, and that ChatGPT invented a Missouri Senate race that was not on the ballot at all. Is that right?