Reports say generative AI is increasingly being used for promotional copy, work emails and other tasks at U.S. publishers HarperCollins, Simon & Schuster and Hachette. Employees and literary agents are calling for clear policies on its use and for publishers to explain their practices to authors. At Simon & Schuster, employees are also opposing the company’s consideration of a workplace analytics tool.
Strict Rules for Authors’ AI Use, but Adoption Within Publishers
The reporting is based on anonymous interviews with more than 20 people working in the publishing industry. According to employees, at least three of the five largest U.S. publishers use tools such as Claude and ChatGPT to write promotional copy, book jacket descriptions, marketing materials and emails to literary agents.
The publishing industry has also seen cases in which books were canceled after their authors were suspected of using generative AI. Publishers take a strict stance toward authors’ manuscripts while using AI to produce writing that reaches readers. How to explain this distinction to authors and readers is at the heart of the issue.
However, generating a novel’s text and writing promotional copy are not the same thing. Because the uses differ, publishers need to specify exactly what they permit and what they do not; otherwise, people outside the companies cannot judge their policies.
Is Promotional Copy Merely an Administrative Task?
One use frequently cited by employees is writing promotional copy for books. Such text reaches readers on online retail sites and on back covers, and is also used when introducing books to booksellers, journalists, reviewers and others. Though produced behind the scenes of publishing, it serves to connect works with readers.
Simon & Schuster says employees across departments can use a limited range of enterprise AI tools, but are not required to do so. The company prohibits the use of unapproved tools and says its aim is to free up time for creative work with authors.
Hachette also says it supports operational uses that help authors’ books reach more readers. However, it does not support creative uses, including communications with authors and business partners. It remains to be seen how closely that policy aligns with employees’ accounts of how email is actually being written.
How Much of an Unpublished Manuscript Can Be Entered?
Literary agents are concerned that authors’ unpublished manuscripts could be entered into AI tools to produce book descriptions or rejection letters. Literary agents handle contract and rights negotiations with publishers on authors’ behalf. One agent interviewed said they had added a contract clause prohibiting manuscripts from being entered into large language models.
It is important to distinguish where input is sent, whether it is stored and whether it is used for training. The use of cloud-based AI alone does not establish that a manuscript was used to train a model. The information gathered for this report also does not clarify the details of each company’s contractual terms or data-management settings.
At HarperCollins, one account says the legal department rejected an editor’s request for permission to enter a full manuscript into an external AI tool. The existence of concern and the fact that a manuscript was actually shared externally are two separate matters. No information has been presented to substantiate the latter.
When Efficiency Proposals Prompt Employee Backlash
At HarperCollins, employees were reportedly asked to attend regular meetings to explore ways to use Claude after the company introduced licenses for the tool. At Simon & Schuster, employees say an internal contest soliciting ideas for AI use was held in May 2024, with a $10,000 prize for the winner and $5,000 each for two other finalists.
Dissatisfaction with these initiatives is not directed solely at AI’s capabilities. Some employees say staff cuts have increased workloads, and that employees who had been asking for adequate staffing and better compensation now face the additional burden of finding ways to use new tools. However, the scale of staffing reductions and workload increases is based on anonymous accounts and does not constitute statistics showing conditions across each company.
There is a gap between companies’ explanation that AI will free up time for creative work and employees’ sense that they have to use it to compensate for understaffing. Evaluating its adoption requires looking not only at time saved but also at whether employees’ workloads have actually eased.
Consideration of Workplace Analytics Tool Exposes Distrust
At Simon & Schuster, the company’s consideration of a product from Skan AI prompted employee backlash. Employees gathered signatures for an open letter to CEO Greg Greeley, urging the company not to introduce similar software into the workplace. They are concerned that, beyond monitoring work, its ability to identify tasks that could be automated could lead to jobs being replaced.
An AI tool that drafts promotional copy and a tool that analyzes employees’ work have different implications. The dispute over the latter shows that publishers’ use of AI extends beyond writing assistance to concerns about how work is allocated and about job security. However, it has not been established that the current consideration of the tool is intended to lead to staff cuts.
In a memo to employees, Greeley acknowledged that the company is considering Skan AI’s Blueprint, but said it has not decided to use the product or any other productivity-monitoring or business-analytics tools.
Comments