TL;DR: If you’re choosing a ghostwriter now, look for one who works with AI the way my clients already do. Every one of them arrives using it to gather material, check facts, and pressure-test the draft, and a real AI-assisted book project shows what that engagement looks like from inside my practice.
How are ghostwriting clients using AI now?
For a while the open question in ghostwriting was whether clients would stop hiring writers and just use the chatbot themselves. That was the threat everyone watched for. It’s not the thing that happened.
What happened instead is that clients started bringing AI into the project as a tool they use alongside me, and the change wasn’t gradual or partial. It arrived across the whole roster at once.
A writer who claims AI does most of the writing is going to hand you something average.Share on X
A client sends a transcript they cleaned up with AI before our call. Another runs a draft chapter through a model to check whether the argument holds and sends me the notes. A third uses it to pull together background research on their own industry so our interview time goes to the parts only they know. None of them asked permission. They simply assumed I’d be fine working this way, because to them it’s just how work gets done now. They can use ChatGPT all they want. I won’t take its text verbatim, because then they’d be hiring ChatGPT and the book would read like it. Its output can look perfect one day and like crap a couple of days later.
It’s a relief that they don’t ask. I use AI every day as a digital assistant that helps me write: transcripts, summaries, first-pass outlines. I still do the writing. A client who shows up already working that way speaks my language, and I’d much sooner work with them than with someone who thinks the tool is either magic or cheating.
Are all ghostwriting clients using AI?
The uniformity stands out. When a new tool shows up in a field, the usual pattern is that one client in ten tries it, then a few more, then it spreads over a year or two. This looked more like everyone arriving at the same understanding in the same season. The executive, the founder, the coach, the first-time memoirist all turned up already using AI in their own work and already expecting the person they hired to be comfortable with it.
That tells you the whole curve moved. Using AI to gather, check, and draft has become the default assumption a professional brings to any knowledge work, and a book project is knowledge work. The client assumes their ghostwriter uses AI, the way they’d assume the writer uses a word processor.
What did client AI use change about ghostwriting?
The old model of ghostwriting was closer to a handoff. The client gave the writer material and access, the writer disappeared for months, and a manuscript came back. In that model the client was a source, then an approver. The work in between belonged to the writer alone.
That’s not the model clients want now. They want to be in the loop, and they want the loop to include the tools they already use. They’ll draft a section badly themselves and hand it over for the writer to fix. They’ll challenge a chapter using a model’s critique and expect the writer to engage with it instead of wave it off.
More and more clients come to me with a book they tried to write with AI that turned out badly, and fixing those books has become a big market. I don’t lecture anyone about it. The failed draft still tells me what they care about, even when almost every sentence in it has to go. The engagement has shifted from handoff to collaboration. A ghostwriter who treats the client’s AI use as interference instead of part of the process is fighting the way the client wants to work.
Ghostwriting has moved from a handoff to a collaboration that includes the tools clients already use.
What do clients expect from a ghostwriter now that they use AI?
The expectation is fluency. The client isn’t impressed that a writer uses AI, any more than they’d be impressed that a writer uses spellcheck. What they expect is that the writer knows where the tool helps and where it ruins the work, and can explain the line clearly.
That line is the actual expertise now. A client who uses AI loosely will produce flat, average prose and may not know why it reads wrong. The ghostwriter’s job includes knowing that the connective tissue, the research, and the transcript cleanup are safe for the machine, while the voice, the argument, and the stories aren’t. The client wants a collaborator who holds that line on the parts the book depends on. That rules out the purist who refuses the tool and the hobbyist who lets it write the book.
In my own practice the line looks like this. I record every call, in Zoom for remote interviews and on a tablet in person, and I say so up front and in the contract. AI helps me turn each session’s transcript into a summary document that goes back to the client. The voice, the argument and the stories stay with me, and nobody should hire a writer who can’t describe their own version of that line.
A client who uses AI loosely will produce flat, average prose and may not know why it reads wrong.
The cornerstone piece on whether AI can write your book lays out where that line sits, and it’s the same line the best clients are already feeling their way toward on their own.
AI is a tool. Treat it like one and you’ll do fine; let it do your work for you and you’ll have problems. My 40-plus writing handbooks use AI prompts to help writers build a better protagonist and plot, and none of them ask the machine to write the book.
My take
Choose a ghostwriter who can tell you exactly where the AI line sits in their work. Research, transcripts and connective drafts are fair game for the machine, and the voice, the argument and the stories stay human. Skip the purist who refuses the tool, and skip the hobbyist who lets it write your book. If you tried an AI draft and it failed, bring it anyway, because it still shows what you care about. A ghostwriter worth hiring will read it and tell you which parts the interviews have to replace.
What does client AI use mean for choosing a ghostwriter?
Ask the writer how they work with AI.
The answer tells you most of what you need to know. A writer who recoils from the question, or who treats any AI use as cheating, is going to be friction against the way you already work, and is also likely to be slower and more expensive for no added quality on the parts that matter. A writer who claims AI does most of the writing is going to hand you something average.
I know an AI specialist who wrote his book with ChatGPT, no editing at all, and it’s getting positive reviews and doing exactly what he wanted. More power to him. Whether the reader stays engaged is the only test I care about, and if they don’t, it’s just another crappy book no matter how it was written.
The writer you want says what the machine does in their process, what stays in their hands every time, and why.
| Ghostwriter type | Approach to AI | What you get |
|---|---|---|
| The purist | Recoils from the question or treats any AI use as cheating | Friction, slower and pricier work, no added quality |
| The hobbyist | Claims AI does most of the writing | Something average |
| The fluent collaborator | Machine handles research, transcripts, connective tissue; voice, argument, stories stay human | A book that does what you wanted it to do |
A ghostwriter who still treats AI as a scandal is working in a market that’s gone, and writers who bill clients for that nostalgia are fleecing them. The ones worth hiring can tell you exactly where the machine stops and the human has to take over, and they’ll produce a book that does what you wanted it to do.
AI helps clients gather material, and a ghostwriter turns it into a book. The Ghostwriting Advantage explains how to get your book written for you.
Frequently Asked Questions
The book on this: The Ghostwriting Advantage is 254 pages on what the engagement consists of. That’s what changes when the client arrives already using the tools.
