These Aren’t the Books You’re Looking For

The Chicago Sun-Times and Philadelphia Inquirer just published a summer section complete with a reading list.

You know the kind. A blurb-y roundup of beach reads, literary fiction and other picks to help readers sift through the shelves at their bookstore or on their Amazon feed.

The only problem is that the majority of the list’s titles don’t exist.

Brit Bennett’s Hurricane Season, exploring family tensions as a Cat 5 storm nears? Not real. (Though there is a real Hurricane Season novel, by Fernanda Melchor.) Boiling Point by Rebecca Makkai, spotlighting a climate activist? Nope. Here’s Pachinko author Min Jin Lee on the list’s inclusion of “her” Nightshade Market: “I have not written and will not be writing a novel called ‘Nightshade Market.’ Thank you,” she posted on X.

“THE CHICAGO SUN-TIMES HAS NAMED ZERO SAINTS THE BOOK OF THE SUMMER. They didn’t, but that doesn’t matter,” author Gabino Iglesias quipped on Bluesky. (It’s in fact his real book!)

The Sun-Times apologized and blamed its circulation department, which used King Features’ list in the paper’s “Heat Index” insert. The Inquirer also distanced itself from the list in a statement. Turns out a freelancer used AI to generate the list as well as other content in the section.

“It was just a really bad error on my part,” freelancer Marco Buscaglia told the New York Times. “Huge mistake on my part … It’s on me 100 percent,” he told NPR.

That’s probably it for his career. (Maybe.) But this messy little episode is emblematic of far more than AI’s propensity for inaccuracies.

These major papers likely relied on a service to provide summer books coverage because so few still employ critics on staff, have dedicated books sections, or even publish many book reviews. Sure, there’s Goodreads and BookTok, but those recommendations just don’t have the same depth as a review in the New York Times.

That this entire list got produced, formatted and printed in major metropolitan newspapers also speaks to suffocating staff cuts, particularly copy editors. I spent decades in assorted newsrooms in the late 1980s and 1990s, and I can’t count the number of times copy editors saved us from printing mistakes simply by asking “Where did you get this?”

Even a cursory Google search to find cover art would have saved face in addition to ensuring that actual books got a nod in the list. Did no one at the paper, presumably fans of the written word, wonder why they’d never heard of these supposedly new books from bestselling authors like The Martian’s Andy Weir or Pulitzer winner Percival Everett?

If you’re still reading newspapers – admittedly, a shrinking demographic these days – you probably are reading at least some AI-generated content. It’s easier and cheaper than employing actual humans.

It’s bad for all journalism, in my opinion, but definitely bad for book reviewing or recommendations. Anyone who’s done one of these roundups – and I’ve done dozens in the decades I’ve spent writing about books both full-time and on a freelance basis – knows how hard it is. Winnowing a book down to three or four sentences is a challenge. Contemplating the list as a whole is another task, ensuring that genres, backgrounds, and lesser-known authors all have a place in the list. To me, a key part of a roundup is to avoid a cookie-cutter reprise of the national best-sellers people can find without much effort.

My suspicion is that the freelancer knew how much time it would take to do that roundup well, calculated how much (or how little) he’d be getting paid, and relied on AI to do the heavy lifting. And time-strapped staff at the syndicate and the newspapers figured “It’s fine,” if they even thought about it at all. To make matters worse, authors are also battling the unauthorized use of their (real) copyrighted work to “train” AI platforms.

Sure, it’s just one list. But it’s a perfect storm of problems with AI.