Big Publishers Are Suing AI Companies While Quietly Using Claude and ChatGPT to Pitch, Market, and Cover Books

In public, America's biggest book publishers have spent the year at war with generative AI. They sued Google for copyright infringement, the Big Five jointly sued Meta in May, and they have canceled books by authors suspected of leaning on chatbots.

Inside the buildings, the picture looks different. At least three of the Big Five, HarperCollins, Simon & Schuster and Hachette, have quietly brought Claude and ChatGPT into daily work without public disclosure or author consent, more than two dozen workers told WIRED on condition of anonymity.

According to those workers, the tools write publicity and back cover copy, design cover art, generate marketing materials and draft emails to literary agents, often at executives' direction. Publicity copy is one of the most common uses. At least one HarperCollins division has gone "all in," down to AI-generated marketing videos.

Licenses first, plans later

HarperCollins leaders bought Claude licenses from Anthropic months ago, staffers say, "but very quickly realized they didn't know how to implement them." Dozens of employees were "voluntold" to become "AI Champions" in monthly brainstorming sessions. The company also bought ChatGPT and Jasper, a platform that promises "to run end-to-end marketing workflows."

When employees raised legal, ethical and environmental concerns at one session this spring, the leader waved them off as "the cost of doing business." Soon after, the company portal added a line saying AI's environmental impact is "a small part of most individuals' total digital carbon footprint."

"It was such a slap in the face," one employee said. "We've been making valid business cases for years to be paid a healthy salary, and instead they spent that money on software that most of us don't really want." The average entry level salary at the Big Five and Scholastic in New York was $47,583 in 2023.

At Simon & Schuster, staff sat through a workshop with OpenAI representatives. "The person leading the workshop genuinely had no idea what we did," one attendee said. A suggestion from the session: use ChatGPT to write to agents. In May 2024 the house ran an internal contest for the best AI use cases, with a $10,000 prize, before rolling back some of the push after complaints at town halls.

Agents have noticed. Some suspected editors were using AI for rejection letters, which "all had the same classic signs of AI writing, like 'not this, but that' and the rule of three," one agent said. Employees at all three houses believe the suspicion is right.

The irony runs deeper. Agents now add contract clauses saying "this material can't be run through an LLM," meant to protect authors from tech companies but binding on publishers too. At HarperCollins, editors asked to paste full manuscripts into an open LLM, and legal told them no.

Skeleton crews and a surveillance fight

Workers tie the shortcuts to layoffs. "It's skeleton crews everywhere you look," one said, describing large teams cut to five or six people handling double or triple the books. "I don't judge them for that. I judge the company for shoving the shortcuts down our throats and saying, 'Do this or we fall behind.'"

The sharpest revolt is at Simon & Schuster, owned since 2023 by KKR. The company trialed Skan AI, a tool that promises leaders "the evidence base they need to reduce costs, prioritize automation, and meet board-level transformation mandates," at KKR's apparent urging. Employees signed an open letter to CEO Greg Greeley voicing "strong opposition" and asking him never to bring Skan AI "or other similar software into S&S workflows." They fear surveillance and "automation discovery" tools built to shadow and then replace them.

The day New York magazine's Book Gossip newsletter reported the letter, October 6, Greeley sent a memo confirming the company had explored Skan's Blueprint product but "no decision has been made," adding that he had not seen the materials before concerns reached him. The AI detector Pangram indicates at least some of that memo was itself AI written.

Simon & Schuster global CMO Wibke Grutjen said staff have access to "a limited number of enterprise-level AI tools," that use isn't mandatory, and that "the goal is always to free up more time for our employees to spend on deeper focus and creative work." Hachette said it supports AI "for operational purposes" but not "creative uses of AI, including communicating with our authors and partners." HarperCollins did not respond.

Creative workers everywhere are running into the same wall, from authors to cartoonists whose real names ChatGPT keeps signing to fake work. Agent Emma Dries of Triangle House Literary wants a simple fix: "We need publishers to be clear about what is being used and when."

Nobody has told staff what to say to authors, one HarperCollins worker said, "because I think they know it would start a shitstorm." Dan Sinykin, author of Big Fiction, reads the moment plainly: the C-suites "are trying to see what they can get away with before there's a revolt."

Some offers on this page may be paid placements or contain affiliate links.