There’s a reason the mention of AI, particularly in creative spaces, gets a bit of an eyeroll. Actually there’s several. It’s trained on stolen content for starters, robbing real artists and writers of credit and income. Furthermore, it’s often just pretty bad, especially when it comes to factual articles. Language models like ChatGPT are known to hallucinate pretty badly, and this has led to real outlets like the Chicago Sun-Times printing a summer reading list full of fake books.
Several outlets have covered the story, such as Arstechnica and The Verge, and of course now I’m doing it here. It could be that we are somewhat motivated to point out when AI stuffs up in the writing space, considering people seem to want to keep giving our jobs to it. But it was 404, which is a paywalled publication, who found the origins of this fake list that made its way into a few publications.
The Chicago Sun-Times made a post on Bluesky, which rather passes the buck on the situation. “We are looking into how this made it into print as we speak,” it reads, adding “It is not editorial content and was not created by, or approved by, the Sun-Times newsroom. We value your trust in our reporting and take this very seriously. More info will be provided soon.”
It turns out the list was bought from a partner of the publications, and was found to come from the media conglomerate Hearst. The listicle features some real books but it’s also plagued by some that don’t exist, credited to both real and fabricated authors. It even points to non-existent blog posts, and is generally just a bout of confusion. Especially for anyone actually trying to get their hands on any of these recommended summer reads.
The byline on the list belongs to a Marco Buscaglia, who 404 managed to track down. Initially Buscaglia admitted to using AI in their work, but clarified that they always check it for errors. “This time, I did not and I can’t believe I missed it because it’s so obvious. No excuses,” he told 404. “On me 100 percent and I’m completely embarrassed.”
This isn’t unique. There were other similar articles found, without bylines, that had blatantly fabricated information with quotes from fake people. One about “Summer food trends” had expert quotes from a doctor that doesn’t exist, as well as some that were never said by people who do. It’s likely this is only the tip of the iceberg when it comes to published hallucinating AI content.
It comes at a time when budget cuts are causing lots of publications to turn to AI content to save money, but it’s definitely a case of you get what you pay for. The sad truth is that there’s far less money for writers of good, well researched, and well written content out there then there used to be. I say this as someone who’s watched publication after publication in my industry close, leaving talented and dedicated journalists without work.
It’s another reminder that we have to be ever careful in what we read, both in print and online. It’s also a reminder for those who use AI that these things are a tool. They need to be used carefully and properly, with the correct oversight. It’s increasingly important to take all your information with a healthy dose of sceptism no matter what side of the readership you’re on.