WIRED: three Big Five publishers quietly use AI for book copy, covers and agent emails - without disclosure or author consent

What happened

On October 9, 2026, WIRED published an investigation, based on interviews with more than two dozen publishing-industry workers (most speaking anonymously for fear of losing their jobs), reporting that at least three of the Big Five US publishers - HarperCollins, Simon & Schuster, and Hachette - have incorporated generative-AI tools into everyday publishing work without public disclosure or author consent.

What the reporting describes

• Staff at the three houses described using ChatGPT, Claude and (at HarperCollins) Jasper to write emails to literary agents, publicity and back-cover copy, cover art assets, and marketing videos - often, workers said, at the direction of executives.
• At HarperCollins, senior leaders purchased Claude licenses from Anthropic and drafted dozens of staff as "AI Champions" to find uses in monthly brainstorming sessions; one division has gone "all in," including AI-generated marketing videos, according to a current employee.
• At Simon & Schuster, OpenAI representatives ran a staff workshop suggesting ChatGPT for composing emails to literary agents. Agents reported comparing rejection letters from the same editors and spotting recurring machine-prose markers - the "not this, but that" contrast construction and lists of three - and employees at all three houses told WIRED they believe the agents' suspicions are right.
• Employees said the widely publicized cases of books canceled over suspected LLM use - including Hachette's horror novel Shy Girl and Macmillan's Call Me, I'll Hide the Body - are only a small part of the related incidents.
• Agents told WIRED they have begun adding contract clauses saying material may not be run through an LLM. Staffers' central worry is whether author manuscripts are being pasted into outside LLMs, and that publishers should disclose what is used and when. The reporting also cites an AI-detection firm's claim that at least part of the Simon & Schuster CEO's memo answering an employee open letter was written by AI.

Company responses

Hachette said it supports operational uses of AI that help books reach more readers, but not creative uses, including communicating with authors and partners. Simon & Schuster said employees have access to a limited set of approved enterprise AI tools and that use is not mandatory. HarperCollins did not respond to WIRED's requests for comment. Days earlier, employees had protested Simon & Schuster's potential adoption of Skan AI's productivity tools; CEO Greg Greeley told staff no decision had been made to adopt them.

Why it matters to creators

Publishing has publicly positioned as one of the loudest opponents of generative AI - suing AI companies over copyright and canceling books whose authors were suspected of using LLMs - while, per this reporting, quietly deploying the same tools on and around authors' work without disclosure or consent. Authors currently have no reliable way to know whether their manuscripts, covers, or marketing materials involved AI processing.