AI-created gaffe embarrasses US mainstream newspapers with phony experts and phony books

Two of America's oldest publications — the Chicago Sun-Times and Philadelphia Inquirer — had to pull content from a seasonal supplement after it was found to contain doctored quotes from non-existent experts and bogus book titles, all created with artificial intelligence, the Washington Post reported.

The mistake was made in the "Heat Index" special section, a multi-page summer-themed syndicated supplement produced by King Features, owned by Hearst. The content was tips, trends, and lifestyle stories for summer, and appeared in print editions of both papers last week.

Fabricated quotes and fictional books set warning signs

The alarm was initially sounded by 404 Media, a technology publication, when the content started going viral on the web and attracted the attention of readers. Among the issues: quotations put into the mouths of imaginary experts, unsubstantiated claims to author interviews, and a summer reading list filled with fake books by big-name authors.

For example, its reading list named Isabel Allende's Tidewater Dreams and Andy Weir's The Last Algorithm — books which do not exist. It also mentioned phony titles by Brit Bennett, Min Jin Lee, and Taylor Jenkins Reid. Furthermore, the insert quoted van-life writer Brianna Madia allegedly interviewing Outside Magazine in 2023 regarding hammocks — a quote that is nowhere to be found in her confirmed interviews.

Newspapers respond with internal reviews and retractions

Victor Lim of Chicago Public Media described the incident as "unacceptable" and promised to re-examine collaborations and content vetting procedures: "We value our readers' trust in our reporting and take this very seriously."

Philadelphia Inquirer publisher and CEO Lisa Hughes affirmed that the section was pulled from the newspaper's online edition. "Having artificial intelligence generate content, as apparently occurred with some of the Heat Index pieces, is a breach of our own internal standards and a serious offense," she said.

Author confesses to AI use, bypassing fact-checks

Much of the content was written by freelance writer Marco Buscaglia, who is located in Chicago, and who, in an interview with The Washington Post, admitted he utilized generative AI tools — presumably ChatGPT or Claude — to assist in gathering content for the supplement. He added that he wished he had cross-checked the information provided by AI.

"There's no excuse," Buscaglia confessed. "I do check things out, but in this case, I totally missed it."

He went on, "I feel worse that somehow the papers get linked to that. I didn't know who would publish [the content]."

AI's increasing role in journalism under review

This mistake is against a background of increasing alarm over the potential misuse of generative AI in reporting. While AI chatbots may help with writing and idea generation, they frequently hallucinate — fabricating facts, quotes, and sources with jarring authority. If these productions are not reliably fact-checked, they can slip into published material, undermining the credibility of the media.

The media's acceptance of AI needs to be balanced with strong editorial control, critics contend. The price of cheating with AI content is higher than a retraction — it's a hit to the credibility of journalism at a time when public confidence is already low, as this episode shows.