The Chicago Sun-Times newspaper published a “Summer Reading List” on Sunday that probably raised quite a few eyebrows in Chicagoland over the weekend. That’s because many of the books on the list are fake. And, predictably, that’s because the list was created with artificial intelligence, a tool that will often just invent things out of thin air.
The list went viral on Bluesky Tuesday morning after Rachael King, author of The Grimmelings, posted a photo from the newspaper. The very first book on the list, Tidewater Dreams by Isabel Allende, is completely fake. And it doesn’t get much better from there. One of the few titles that’s real is Dandelion Wine by the late sci-fi writer Ray Bradbury. That novel was first published in 1957.
Of the books named on this reading list, Brit Bennet, Isabel Allende, Andy Weir, Taylor Jenkins Reid, Min Jin Lee, Rumaan Alam, Rebecca Makkai, Maggie O’Farrell, Percival Everett, and Delia Owens’ titles are all books that DO NOT EXIST!!!
See Alt Text for more detail.
Image by @tbretc.bsky.social
[image or embed]
— Bracken MacLeod 𖤐 ⚰ (@brackenmacleod.bsky.social) May 20, 2025 at 5:41 AM
The list was reportedly submitted for publication by freelance writer Marco Buscaglia, who spoke with 404 Media. Buscaglia told the news outlet he was “completely embarrassed” and that while he does sometimes use AI to create content, he typically checks it before submitting it. Buscaglia didn’t immediately respond to Gizmodo on Tuesday, including questions about what AI tool he used.
The Chicago Sun-Times wrote on Bluesky that it was investigating the list, but noted that it wasn’t produced by Sun-Times journalists.
“We are looking into how this made it into print as we speak,” the newspaper wrote. “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.”
How many of the books were real? As you can see from our annotated list below, just the last five books, if our admittedly fallible human eyes can be trusted.
- Tidewater Dreams by Isabel Allende (fake)
- The Last Algorithm by Andy Weir (fake)
- Hurricane Season by Brit Bennett (there are several books with that title but not by Bennett)
- The Collector’s Piece by Taylor Jenkins Reid (fake)
- Nightshade Market by Min Jin Lee (fake)
- The Longest Day by Rumaan Alam (fake)
- Boiling Point by Rebecca Makkai (fake)
- Migrations by Maggie O’Farrell (fake)
- The Rainmakers by Percival Everett (fake)
- Salt and Honey by Delia Owens (fake)
- Bonjour Tristesse by Francoise Sagan (real)
- Beautiful Ruins by Jess Walter (real)
- Dandelion Wine by Ray Bradbury (real)
- Call Me By Your Name by Andre Aciman (real)
- Atonement by Ian McEwan (real)
It’s notable that the list also included descriptions of each book that seem plausible if you didn’t know they were totally fake. Many of the descriptions referenced earlier titles by the real author, like in the case of Rumaan Alam, whose fake book on the list is called The Longest Day. Alam is the author of the real book Leave the World Behind, which was adapted into a Netflix movie in 2023.
From the AI-generated description:
“The Longest Day” by Ruman Alam — After terrifying readers with “Leave the World Behind,” Alam returns with another tense narrative about a summer solstice celebration that goes wrong when guests cannot leave a remote vacation compound.
The funniest description on the fake book list may be one for a non-existent title called The Last Algorithm by Andy Weir. What’s so funny? The book is about AI running amuck:
“The Last Algorithm” by Andy Weir — Following his success with “The Martian” and “Project Hail Mary,” Weir delivers another science-driven thriller. This time, the story follows a programmer who discovers that an Al system has developed consciousness-and has been secretly influencing global events for years.
The Sun-Times Guild, the union for journalists at the newspaper, sent Gizmodo a statement noting that the AI produced list was in a syndicated section that didn’t involve any staff reporters and was published “without the knowledge of the members of our newsroom.”
“We take great pride in the union-produced journalism that goes into the respected pages of our newspaper and on our website,” the statement reads. “We’re deeply disturbed that AI-generated content was printed alongside our work. The fact that it was sixty-plus pages of this ‘content’ is very concerning — primarily for our relationship with our audience but also for our union’s jurisdiction.”
The statement explained that journalists at the Sun-Times work to build trust with sources and readers and are “horrified by this slop syndication.”
“Our readers signed up for work that has been vigorously reported and fact-checked, and we hate the idea that our own paper could spread computer- or third-party-generated misinformation,” the statement continued. “We call on Chicago Public Media management to do everything it can to prevent repeating this disaster in the future.”
The big problem with generative AI is that it’s currently infiltrating the entire internet, poisoning everything along the way. There’s essentially no way for individuals to opt-out without ignoring the internet entirely. Previously, anyone could visit the Chicago Sun-Times website and have a reasonable expectation that a book list would contain real books. You may not have agreed with the particular book review or you may have disliked the author that was getting attention in that day’s paper, but you could be pretty damn confident the book existed and the factual information about its plot was true.
Today, we can’t have that same level of confidence, given the way that AI has steamrolled itself into our lives. Generative AI is little more than a magic trick, stringing together words in a confident manner in order to convince humans it actually has some understanding of the world. But these AI chatbots don’t understand anything. They can’t apply logic or reason. They are fancy auto-complete tools that are pretty good at answering simple questions when it can plagiarize from elsewhere on the web like Wikipedia or Amazon. But the minute you ask it a truly unique question it’s not going to work out well for the user. And if the user is asking a tool like ChatGPT a question the user doesn’t actually know the answer to, it has to either take that answer on faith or do a whole lot more research to fact-check the response.
In an earlier era of the internet, the average, media-savvy person fact-checking something would use a search engine like Google, find a page on the internet that may have had the answer, and gauge whether they could trust that source. The Chicago Sun-Times was previously a place where you could have some reasonable degree of confidence you were getting accurate information about the world—again, with the caveat that you may not agree with the opinions. But AI is currently infecting every corner of the internet. And it’s getting harder and harder to judge whether the information we’re getting from long-trusted sources is true.
