Big Five Publishers Quietly Use AI While Fighting It Publicly

Three of the Big Five US book publishers have quietly folded generative AI into day-to-day operations while publicly battling the technology in court and on social media, according to more than two dozen publishing employees who spoke to WIRED on condition of anonymity.

At HarperCollins, Simon & Schuster, and Hachette, staff now use large language models including Anthropic's Claude and OpenAI's ChatGPT to draft emails to literary agents, write back-cover and publicity copy, generate cover art, and produce marketing materials. None of the deployments came with public disclosure or author consent.

The public posture says the opposite. American publishers spent 2024 suing OpenAI and Google over copyright and canceling books by authors accused of leaning on LLMs. Hachette dropped a horror novel called "Shy Girl"; Macmillan canceled a crime thriller called "Call Me, I'll Hide the Body." Those cancellations are merely "the tip of the iceberg," one editor tells WIRED. "Every single imprint has a story of someone very esteemed who had a book canceled because of blatant AI use."

What is actually happening inside the publishers?

HarperCollins bought Claude licenses from Anthropic "months ago," a current employee says. Senior leaders "very quickly realized they didn't know how to implement them." The company then "voluntold" dozens of junior and senior staffers to serve as "AI Champions" tasked with brainstorming monthly use cases.

The pushback started fast. At a recent session, when employees raised concerns about legal, ethical, and environmental fallout, a meeting leader dismissed them as "the cost of doing business," according to someone who attended. The HarperCollins employee portal soon added language arguing that AI's environmental footprint is "a small part of most individuals' total digital carbon footprint."

Some staffers felt blindsided. "It was such a slap in the face," one employee says. "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 New York City offices of the Big Five and Scholastic was $47,583 in 2023.

HarperCollins has also licensed ChatGPT and Jasper, an agentic platform that markets itself as a way to "run end-to-end marketing workflows." One HarperCollins division has gone "all in," producing AI-generated marketing videos. HarperCollins did not respond to multiple requests for comment.

What are they using the tools for?

Publicity copy tops the list. Publicists shape how books appear on Amazon, Goodreads, and back covers; they pitch reviewers and booksellers with that copy. "Using AI to generate publicity copy is short-sighted," a Big Five publicist says. "Publicists have a duty to engage in good faith with readers and our colleagues in the media, who I'm sure have a low threshold for tolerating AI-generated pitches."

Meg Reid, publisher of the independent Hub City Press, sees the gap opening further. LLM-written publicity "feels pretty inevitable" for the Big Five, she says. "That's one place we're probably going to see a bigger shift between corporate publishers looking for efficiencies and smaller publishers insisting that the humans involved are the whole point."

Agents have noticed too. Earlier this year, literary agents began suspecting specific Big Five editors of using AI for rejection letters. "Among colleagues, we compared emails on different books from the same editors, and they all had the same classic signs of AI writing, like 'not this, but that' and the rule of three," one agent says. Employees at Simon & Schuster, HarperCollins, and Hachette tell WIRED they believe the agents' suspicions were correct.

Staff at Simon & Schuster sat through a workshop with OpenAI representatives months ago. "The person leading the workshop genuinely had no idea what we did," an attendee says. One of OpenAI's actual recommendations for the book trade: compose emails to literary agents. That practice has spread across editorial teams at HarperCollins and Hachette.

A Hachette spokesperson said the company "support[s] the use of AI tools for operational purposes, when they help our authors' books reach more readers," but does not support "creative uses of AI, including communicating with our authors and partners." Wibke Grutjen, global chief marketing and communications officer at Simon & Schuster, said "employees across all functions have access to a limited number of enterprise-level AI tools," though their use is not mandatory. The company prohibits any tool that has not been "vetted and approved."

How far does the experimentation go?

Simon & Schuster ran an internal contest in May 2024 asking employees to propose AI use cases. The winner got $10,000; two runners-up got $5,000 each. The company has since "rolled back some of the AI-pushing because employees kept complaining about it in town halls," one employee says. AI-generated assets are also creeping into cover art. Some foreign presses distributed by Simon & Schuster openly acknowledge AI use when sharing designs.

Hachette employees are bracing for tighter AI rules under new chief information officer Carl Hixson. Hixson previously served as chief technology officer at SRMG, a Saudi state-backed media company that has publicly embraced agentic AI.

What do authors and agents fear most?

Agents fear publishers might paste unpublished manuscripts into LLMs to draft rejection letters or publicity descriptions. "We've added AI clauses to our contracts saying 'this material can't be run through an LLM,'" one agent says. "Ostensibly, that was meant to protect authors from tech companies, but it also applies to the company buying and publishing the book. If they're not using a closed-loop model, they could be feeding an LLM with an author's unpublished writing."

A closed-loop model runs locally on a publisher's hardware, with no cloud connection. Most consumer tools, including ChatGPT and Google Gemini, are open-loop by default. One HarperCollins source says multiple editors asked managers if they could paste full manuscripts into an open-loop LLM and were told no by the company's legal team.

What is Skan AI, and why is it different?

In the past few weeks, word spread at Simon & Schuster that the company was running an experimental trial with Skan AI. The move came "at the apparent urging of KKR," the investment firm that acquired Simon & Schuster in 2023. Skan AI's website describes its product as giving leaders "the evidence base they need to reduce costs, prioritize automation, and meet board-level transformation mandates."

Simon & Schuster employees responded with an open letter to CEO Greg Greeley expressing "strong opposition" and seeking a commitment to "not bring Skan AI or other similar software into S&S workflows, either now or in the future." Beyond surveillance, workers pointed to the platform's "automation discovery" tools, which can map workflows and identify roles for replacement.

New York magazine's Book Gossip newsletter reported the open letter on October 6. That same day, Greeley circulated a memo confirming that Simon & Schuster had explored Skan AI's Blueprint product. "No decision has been made to use Blueprint or any other productivity-monitoring or workflow-analysis tool," the memo said. Greeley added that he "had not seen the materials relating to this early-stage exploration" before concerns were raised the day before.

Pangram, the AI-detection service that has become a default check for suspicious prose, indicated that AI generated at least some of Greeley's memo.

Why are employees going along with this?

Several staffers say they adopted AI tools reluctantly, after layoffs thinned teams and tripled workloads. "It's skeleton crews everywhere you look," one employee says. "Large teams have been reduced to just five or six people, give or take, and we're doing double or triple the number of books. No wonder people are taking shortcuts. 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.'"

What do publishers say, in public?

Hachette's official line stops at operational use. The spokesperson said the company supports AI "when they help our authors' books reach more readers" but draws the line at "creative uses." Grutjen, at Simon & Schuster, said the goal is "to free up more time for our employees to spend on deeper focus and creative work on behalf of and alongside our authors."

Almost two years ago at the UBS Global Media and Communications Conference, HarperCollins CEO Brian Murray told investors that AI presents "more opportunities than risks" for publishers and pointed to audiobooks and translations as growth areas. Editors say they have since seen LLM-translated manuscripts bearing AI-use disclosures at the Frankfurt and London book fairs.

Michael Reynolds, executive publisher at Europa Editions, told booksellers at the New Voices New Rooms conference in Baltimore in August that he welcomed AI's potential in translation. Attendees say he "seemed baffled by critiques of it." Europa Editions, known for publishing Elena Ferrante's Neapolitan Quartet, already omits human translator names from its covers.

What do industry watchers want next?

Most agents and editors asked for one thing: clear guidelines and disclosures. "We need publishers to be clear about what is being used and when," says Emma Dries, an agent at Triangle House Literary and former editorial staffer at HarperCollins and Penguin Random House. "Short of these tools being banned, which would frankly be my preference, everybody should be communicating more."

A current HarperCollins staffer offers a colder read of management silence: there has been no direction "about what to say to authors or agents about any of this, because I think they know it would start a shitstorm."

Dan Sinykin, author of "Big Fiction: How Conglomeration Changed the Publishing Industry and American Literature," sees the gap as structural. "The C-suites at Big Five publishing companies have more in common with the C-suites at other major American companies than they do with the editors and publicists and translators who work for them," he says. "They're always looking for ways to maximize profit, and right now they're trying to see what they can get away with before there's a revolt."

The next flashpoint sits at Simon & Schuster, where contracts now explicitly bar LLM ingestion of unpublished manuscripts. Whether those clauses hold against internal pressure to feed in-house manuscripts into productivity software will test whether the industry's first AI revolt can produce more than a memo.