A Credibility Problem, Printed on Sixty-Four Pages
The Chicago Sun-Times recently mailed subscribers a sixty-four page, full-color special insert branded with the newspaper’s logo. To anyone who has spent time reading text produced by large language models, the material reads as unmistakable "AI slop." It is replete with an oily sheen of enthusiasm and an over-reliance on adjectives, but the more serious issue is the apparent fabrication of quotations and sources throughout the publication.
One reader took the time to check the claims, and found that the insert does not survive even a cursory fact-check. A complaint letter to the publisher detailed several suspicious passages that fail to match any verifiable source material.
Fabricated Experts and Phantom Books
The insert’s "Summer soundtracks" article cites Dr. Daniel Levitin in an interview with Scientific American, attributing to him a quote about music activating the brain’s reward centers and serving as "powerful memory anchors." A search for the purported phrasing and its substrings returns zero matches on Kagi or Google. While Scientific American did publish an article on January 22, 2001 featuring Levitin, its content bears no resemblance to the quoted text.
A "Nightfall adventures" piece goes further, quoting Dr. Howard Whiteman, a professor at Murray State University. When contacted, Whiteman categorically denied ever saying the words attributed to him:
I have never spoken to the Chicago Sun-Times or any other Chicago paper, and I can’t recall ever saying any of those words to anyone, ever. I write for some regional papers, some of which is online, and I could see an AI putting all of my writing together somehow into those words, but I’ve never actually written about nocturnal behavior per se.
The problems extend to non-existent texts. A summer reading list recommends The Rainmakers by Percival Everett and Salt and Honey by Delia Owens. Both authors exist, but no evidence suggests either book exists. The citation of "hyperlocal ingredient sourcing" and "unexpected savory notes" as the "fastest-growing segments in premium frozen desserts," attributed to the National Ice Cream Retailers Association, also appears nowhere on that organization’s web site.
Failure of Editorial Oversight
Some articles lack a byline, while others are attributed to a single journalist. The insert has no identifiable editorial path that would explain how fabricated quotes from real scientists and recommendations for non-existent books could proceed through production. The publication fails an LLM "sniff test" on essentially every page. The question is whether the named writer was negligent, or whether they, like the experts cited, were also a victim of hallucinated attributions from a statistical model.
