AI journalism mistakes: Live tracker of major mishaps

Four stories cited in Press Gazette round-up of AI going wrong in news publications

AI is being widely used in journalism and can lead to reputation-killing scandals and mistakes if not monitored closely. Here Press Gazette rounds up some of the main examples of where AI has gone wrong.

Most recently, The New York Times published a quote attributed to Canadian opposition leader Pierre Poilievre that was an AI-generated summary of his views, using words he had not said.

The Mississippi Free Press admitted to being the latest news outlet caught out by publishing an AI column written by a fake author.

The non-profit outlet said the journalist did not seem suspicious until they submitted an invoice that did not match their name.

This was the same way that purported freelance journalist Margaux Blanchard was caught out by Wired last year.

There have been numerous similar cases of AI work mistakenly published by major news outlets in the past year as the technology grows more sophisticated. Sometimes the ‘journalists’ are caught out by commissioning editors pre-publication.

There have also been examples of real writers getting caught out using AI in unsanctioned ways. Australian news website Crikey took down a series of articles because a writer had used ChatGPT in the editing process against its strict AI policy.

In this new page, Press Gazette will keep track of such incidents to help publishers to learn from these mistakes. Last updated: May 2026

If you spot anything that we have missed, please email charlotte.tobitt@pressgazette.co.uk.

Skip to:

2026 cases

2025 cases

2023-24 cases

Read more in-depth analysis of AI and journalism with a Press Gazette subscription .

AI in journalism: Every confirmed case of fake authors and AI-generated articles

New York Times politician ‘quote’ was AI-generated summary (May 2026)

The New York Times has issued a correction after publishing what was purportedly a quote from Canadian opposition leader Pierre Poilievre but was in fact an AI-generated summary of his views.

The correction was made on 1 May, more than two weeks after publication.

An editors’ note at the bottom of the article states: "This article was updated after The Times learned that a remark attributed to Pierre Poilievre, the Conservative leader, was in fact an AI-generated summary of his views about Canadian politics that AI rendered as a quotation.

"The reporter should have checked the accuracy of what the AI tool returned.  The article now accurately quotes from a speech delivered by Mr Poilievre in April."

The supposed quote in the original article attributed to a March speech by Poilievre said: "If these turncoats have any shred of integrity left, they should resign their seats tonight and run in a by-election tomorrow. Let the people decide if they want a Liberal representing them, because they sure didn’t vote for one in the general election."

In fact Poilievre had not referred to "turncoats". The real quote that now appears in the article (attributed to an April speech) said: "My personal opinion is that when a member of Parliament goes back on the word they made to their constituents and switches parties, constituents should be able to petition to throw them out.

"That would put the people back in charge of our democracy rather than having dirty backroom Liberal deals by Mark Carney determine our destiny."

Mississippi Free Press AI fake author (April 2026)

The Mississippi Free Press has announced it discovered an opinion column published on 7 April was written using AI and "the ‘author’ was not who they claimed to be".

The column was headlined: "The gig economy is affecting our communities."

The page remains live but the text of the article has been removed and replaced with an editor’s note stating: "This column did not meet MFP’s standards and has been removed."

Kevin Edwards, editor of the Voices opinion section, told readers the author has been "purged from our system" and admitted "the mistake was mine".

He explained: "The AI column submitted by this author didn’t seem out of the ordinary. In fact, it wasn’t until they submitted an invoice that didn’t match their name that I grew suspicious. Not of the column itself, but of the author. Was this person who they claimed to be?

It turns out that, no, they weren’t. I looked back at our email correspondence and checked out the various social media links they had provided in their email signature. All were dead or nonexistent.

"I searched their name with a company listed on their résumé and found an editor who had already gone through the same song-and-dance with the writer, though he figured out the ruse before he published a fake article. On closer inspection, it turns out that the headshot the writer sent us for his bio picture was also generated with AI."

Edwards added that other columns he had recently been sent from new authors also came "from fake authors with...