Major Book Publishers Quietly Adopt AI Tools Across Editorial and Marketing Operations

Major U.S. book publishers have mounted public defenses against unauthorized generative artificial intelligence—filing copyright lawsuits and canceling projects containing machine-generated text. Behind the scenes, however, several top trade houses are quietly deploying enterprise AI software into internal operations without author consent or public disclosures, according to an investigation by Wired .

Interviews with more than two dozen industry workers reveal that major publishers, including HarperCollins, Simon & Schuster, and Hachette, are using large language models like OpenAI's ChatGPT and Anthropic's Claude to draft publicity pitches, create back-cover copy, design cover art, and generate marketing materials. Editorial staff also reported that undisclosed cancellations of manuscripts by established authors due to AI usage have occurred across multiple imprints, extending beyond publicized cases such as Hachette’s Shy Girl or Macmillan’s Call Me, I’ll Hide the Body .

Internal software adoption has accelerated unevenly across houses. HarperCollins acquired enterprise licenses for Claude, ChatGPT, and marketing automation platform Jasper, appointing cross-departmental "AI Champions" to surface practical use cases. Staff pushback regarding the ethical and environmental impacts of AI was characterized by meeting leaders as a cost of doing business, contrasting with industry compensation that averaged an entry-level salary of $47,583 across the Big Five and Scholastic in New York City in 2023.

At Simon & Schuster, which was acquired by private equity firm KKR in 2023, management held an internal contest in May 2024 awarding a $10,000 top prize and two $5,000 runner-up prizes for employee AI use cases. The company also hosted OpenAI-led workshops that proposed using ChatGPT for drafting correspondence to literary agents. More recently, Simon & Schuster employees organized an open letter opposing an experimental trial of Skan AI’s Blueprint process-monitoring tool, prompting CEO Greg Greeley to issue an October 6 memo stating that no final decision to implement the workflow-analysis software had been made.

The growing presence of LLMs has also triggered contractual concerns among literary agents, who reported identifying common AI phrasing in rejection emails and raised alarms about unpublished manuscripts being submitted to open-loop models. While HarperCollins legal counsel instructed staff not to upload full manuscripts into open-loop systems, agents have begun inserting explicit contractual clauses barring publishers from running author texts through AI models.

Publishing leadership has defended the enterprise adoption while outlining intended guardrails. A Hachette spokesperson stated the company supports AI for operational tasks that help titles reach readers, but does not support creative uses or author communications. Simon & Schuster Global Chief Marketing and Communications Officer Wibke Grutjen said access to vetted enterprise tools remains optional and aims to free staff for creative work. Staffers, however, noted that team downsizing and increased title workloads have driven employees toward automated shortcuts.