AI or Not? Debate Erupts Over Authenticity of Gannett's 'Reviewed' Articles, Employees Demand Answers Writers at Reviewed became suspicious of articles that lacked specific author information on platforms like LinkedIn.
By Madeline Garfinkle •
Key Takeaways
- Employees at Gannett are questioning if some Reviewed article were "written" by AI.
- Gannett maintained that the articles were not AI-generated but part of a deal with a marketing firm to boost search-engine traffic.
Opinions expressed by Entrepreneur contributors are their own.
A series of articles on Gannett's product review platform, Reviewed, has ignited skepticism among staffers who began wondering whether the content was generated by artificial intelligence, The Washington Post reported.
Jaime Carrillo, a senior staff writer at Reviewed, told the outlet he and some colleagues were confused about some "vaguely" written reviews, which they suspected were produced by artificial intelligence. He also said that when attempting to find the credited authors, they were nowhere to be found on LinkedIn or other career-professional websites.
"It's gobbledygook compared to the stuff that we put out on a daily basis," Carrillo told WaPo. "None of these robots tested any of these products."
The writing in the articles (many of which have now been removed) was found by The Verge to be "stilted, repetitive, and at times nonsensical."
Related: Business Schools Are Adding AI Education Into The Curriculum
Gannett has previously faced criticism for such content. In August, it conducted an "experiment" using AI to generate sports articles, resulting in repetitive and strange phrases like "he was in hibernation in the fourth quarter." Following a backlash, Gannett suspended the use of AI tools and pledged to reevaluate its methods.
However, on Tuesday, the NewsGuild of New York, the union representing many Reviewed employees, shared screenshots on X, formerly Twitter, of the articles, suggesting Gannett's ongoing use of AI tools to produce content as two screen-shotted articles possess "nearly identical" content.
UPDATE: @Reviewed's shoddy AI consumer reviews are back up again, after @Gannett seemingly shut them down earlier today.
— NewsGuild of New York (@nyguild) October 24, 2023
Here's the link. Which is the worst review on here? https://t.co/oAU3QE5PAE https://t.co/eluoSexhpy pic.twitter.com/sRBsCQ8kgE
Gannett said in a statement that the articles in question were not generated by AI, but were part of a deal with a marketing firm to boost paid search-engine traffic.
"We expect all our vendors to comply with our ethical standards and have been assured by the marketing agency the content was NOT AI-generated," the spokesperson added to WaPo.
Now, affected stories carry a disclaimer. For example, an article titled "Best waist lamp of 2023," which has been removed from the Reviewed website, (but the Internet is forever), held a message at the top: "These pages are published as a partnership between Reviewed and ASR Group Holdings, a leading digital marketing company."
However, many staff writers at Reviewed are still unhappy about the situation, and the union expressed to The Verge its intent to address the issue during the upcoming bargaining sessions, demanding the removal of the articles and a "formal apology" from management.
"It's an attempt to undermine and replace members of the union whether they're using AI, subcontractors of a marketing firm, or some combination of both," the unionized workers said in an email statement to the outlet.
Related: Meta Decides Not to Release AI That Can Mimic the Voices of Everyone You Know
