Journalists at USA Today suspect that parent company Gannett is using artificial intelligence to generate content for a product review site after mysterious bylines of unknown writers started showing up on articles, according to a report.
Reviewed, the USA Today-owned shopping recommendation site, published several articles last week with names of reporters that other staffers did not recognize, according to the Washington Post.
The stories on the site also contained suspiciously similar content and prose, which the union representing USA Today journalists called “robotic.”
When the NewsGuild of New York, which represents Reviewed employees, researched the identity of the authors online, it found that they “may not even be real” as they didn’t have any profiles on LinkedIn.
Google searches of the authors’ stories also turned up nothing, according to the union, which posted screenshots of two separate articles that contained similar language.
A Washington Post reporter who tracked down email addresses linked to the mystery authors wrote that emails “bounced back with a message indicating there was no such inbox.”
One story reviewing scuba masks was juxtaposed with another article that was devoted to tumbler cups.
“Before buying a scuba mask, there are several important factors to consider,” according to the first Reviewed story.
The tumbler cups review article began: “When shopping for a tumbler, there are many factors to consider.”
Tumbler cup shoppers are urged to “choose a design that fits your style” while would-be divers are told to “choose the lens type that suits your preferences and diving style.”
In a statement to The Post, Gannett denied that it was using AI-generated content for Reviewed.
“The Reviewed content referenced was created by third-party freelancers hired by a marketing agency partner, not AI,” a spokesperson for Reviewed told The Post.
The union also accused Gannett of “promot[ing] totally random products” that were not reviewed by staffers, including health supplements.
Gannett was also accused by the union of promoting the stories as sponsored content on Google “over the articles written by Reviewed’s own staff.”
“The pages were deployed without the accurate affiliate disclaimers and did not meet our editorial standards,” a Reviewed spokesperson told The Post. “Updates have been published.”
The union accused management of using AI in “a transparent attempt … to union-bust by threatening reporters with the loss of their jobs.”
Earlier this month, staffers at Reviewed staged a one-day strike to protest “management’s failure to come to the bargaining table” on a new agreement.
Gannett, which owns more than 200 daily newspapers in 43 states, did not address the union-busting charge in its statement to The Post.
Since merging with GateHouse Media four years ago, Gannett, whose properties include the Arizona Republic, the Austin American-Statesman and the Asbury Park Press, has slashed more than half of its workforce and shut down dozens of local newspapers.
In 2015, shares of Gannett were trading at a record high of $25 per share. As of Friday, the stock price was languishing near $2.30 per share.
Several publishers have come under fire in recent months for using AI to publish stories that turned out to be riddled with errors, including G/O Media.
Axel Springer CEO Mathias Döpfner — who heads the $7.32 billion German parent company which owns American news brands including Politico and Insider — shocked staffers earlier this year when he predicted that AI could eventually replace journalists.
In late January, BuzzFeed’s stock price surged after the company announced plans to expand its use of ChatGPT in order to create content on its website.
CNET, a tech-centric website, has reportedly been using AI to generate content.