WIRED reports Big Five publishers are using AI tools behind the scenes

In short: Several large US book publishers are increasingly using AI tools for day to day publishing work, and some employees are pushing back.

What's going on

WIRED reports that at least three of the “Big Five” US publishers, HarperCollins, Simon and Schuster, and Hachette, have been adding AI tools into their workflows without public disclosure or clear author consent. These tools include chatbots like ChatGPT and Claude, which can write text based on prompts (like an autocomplete that writes whole paragraphs).

Workers told WIRED that AI is being used to draft emails to literary agents, write back cover and publicity copy, and create marketing materials. Some staff also said AI is showing up in cover art and even marketing videos. One common use so far is publicity copy, the short descriptions used on back covers and on sites like Amazon.

At HarperCollins, employees said leaders bought Claude licenses but then asked staff to become “AI Champions” and find ways to use it. Some workers said concerns about legal, ethical, and environmental issues were dismissed in meetings. HarperCollins did not respond to WIRED’s requests for comment.

At Simon and Schuster, staff described an OpenAI workshop and an internal contest with cash prizes for AI use ideas. More recently, employees began signing an open letter opposing a trial of Skan AI, a tool they believe could be used to monitor and replace workers. The CEO later said no decision had been made to use it.

What to watch

Agents and employees say the next flashpoint could be transparency. Some agents are adding contract language to stop manuscripts from being pasted into online AI systems, out of fear unpublished writing could leak. Watch for publishers to publish clearer AI rules, and for authors and agents to demand disclosures about when AI is used.

Source: Wired