AI & LLM Policy
We all know by now that artificial intelligence (AI)—and in particular, large language models (LLMs) such as ChatGPT, Claude, and Gemini—have had profound effects on the way we get our information and exchange ideas. But what none of us can know, or predict, is the future implications of these models for the publishing industry.
I therefore have a strict policy regarding the use of AI, both in my own practice and in terms of the writing and editing projects I take on.
My approach to AI
I support the ethical use of AI; however, I have a fairly narrow window of what I consider ethical in terms of writing and publishing.
Specifically, I support the use of assistive AI—that is, scripts or prompts that automate routine tasks or increase the user’s efficiency. This includes, for example, scheduling appointments, running batch cleanup processes, and generating summaries of articles.
That said, I believe that human thought and discernment are irreplaceable for effective communication, and I am very concerned about the increasing reliance on LLMs for correspondence, content generation, and decision-making.
I do not use generative AI of any kind, including LLMs, in either a personal or professional capacity. I also do not use assistive AI tools, with the slight exception of using macros and similar automated scripts as part of the routine cleanup of Word documents. I have never accepted, and will never accept, any job that would require me to train or improve any kind of generative AI model, including LLMs.
This means that, when you work with me, your work will never be exposed to an LLM.
My policy for client work
I do accept manuscripts from clients that were created partially or completely with generative AI, but only under very specific circumstances.
When I begin work with a new client, a key part of the intake process is exploring the extent to which generative or assistive AI was used in the creation of their work. Whether I accept a project that was created with any kind of AI usage is on a strictly case-by-case basis, and is always subject to the requirements outlined below.
In general, I will never knowingly accept or continue a project where the manuscript is out of compliance with any relevant AI ethics and usage policies. I reserve the right to decline or discontinue any project, at any time, where the generative AI or LLM usage seems to be significantly out of step with the overall purpose or intended audience. I also reserve the right to cancel, at any time, any project where the author’s AI usage turns out to be different than initially disclosed, even if the project is already in progress.
Academic works that fulfill a degree requirement: I do not accept manuscripts of this kind that were created with any amount of generative AI usage. This includes bachelor’s and master’s theses, dissertations, and similar works. I may accept manuscripts of this kind where only assistive AI was used. The manuscript must be in compliance with all of your institution’s relevant policies on AI usage and ethics.
Academic works not created to fulfill a degree requirement: I may accept manuscripts created with partial assistance from generative AI. The use of assistive AI is generally allowed. The manuscript must be in compliance with all of your institution’s relevant policies on AI usage and ethics.
Non-academic works: For my “Generative AI / LLM Clean-Up” service, I accept manuscripts created with any amount of assistance from generative AI. Note that purchasing this service means that you agree to me entirely re-drafting the manuscript, using the original as a basis. For developmental, line, or copy editing, I may accept manuscripts created with partial assistance from generative AI. The use of assistive AI is generally allowed. The manuscript must be in compliance with any relevant policies on AI usage and ethics.