Macmillan Employees Express Concerns Over AI Assistant NOA and Its Impact

NOA was developed by Chaptr, a German software company backed by Holtzbrinck, Macmillan’s parent company. An internal email introduced NOA as an AI chatbot designed to be a secure tool for inputting proprietary information, claiming to operate without sharing or retaining data for model training. However, staff were informed during training that NOA could produce inaccurate information and required human verification.

Despite being encouraged to use NOA for various tasks, including email composition and marketing research, many Macmillan employees found the tool ineffective. One employee described NOA as “terrible” and noted minimal usage among staff. Macmillan later suggested NOA for setting performance review goals, prompting further pushback from employees.

In October 2025, company leadership announced via email that NOA would be discontinued. Following this, employees began exploring the AI tool Google Gemini, but reports indicated that its evaluations were also deemed unhelpful. Kelly Collins, Macmillan’s vice president of communications, acknowledged the use of Gemini for business purposes but clarified that there was no directive to use it for evaluating submission letters from literary agents.

Macmillan is one of the smallest among the Big Five U.S. book publishers and is known for imprints like Farrar, Straus and Giroux and Tor Books. Recent communications from employees across the Big Five suggest that executives are increasingly pressuring staff to utilize AI tools, despite ongoing legal actions from some companies against AI developers for copyright infringements. Notably, several bestselling authors have expressed dissatisfaction, fearing that AI could undermine relationships with publishers and their creative processes.

At Penguin Random House, where ChatGPT Enterprise has been promoted as a productivity tool, employees report a reluctance to adopt it due to its perceived ineffectiveness. Claire von Schilling from PRH stated that the use of AI varies across different publishing groups and emphasizes the need to protect author intellectual property and human creativity.

Authors continue to seek transparency and communication from publishers regarding AI usage, with some expressing that they may consider self-publishing if traditional publishing does not maintain a human-centered approach. The growing trend of using AI tools in publishing has led to mixed reactions, with many authors worried about their jobs and the integrity of the publishing process.