AI will replace the drafting part of ghostwriting, which is most of what a typical retainer pays for. It will not replace the parts founders actually value most: accountability, a strategy session with someone who has done this before, and a human they trust to push back. The ghostwriters who survive will sell those things and use an AI ghostwriter underneath to write the posts.
Yes, AI will replace a lot of ghostwriters. Specifically, the ones whose product is the draft.
If your ghostwriter's main output is eight posts a month written from a short interview, an AI ghostwriter can already produce that, faster, in a voice closer to yours, for a fraction of the price. What AI does not replace is the accountability, the strategic thinking and the human relationship that some ghostwriters provide on top of the writing. Those ghostwriters will do fine. Many will do better, because AI takes away the part of the job that never scaled.
I should say where I'm coming from. I'm building an AI ghostwriting product for founders. Before and while building it, I went through more than 16 ghostwriting agencies and 9 AI-native competitors, and I sat in a lot of sales calls with founders who had hired ghostwriters, fired them, or were deciding. So I'm biased, and I've also seen the model up close from both sides.
What you are really paying a ghostwriter for
Start with the money. From what I've seen, LinkedIn ghostwriters charge anywhere from about $1,500 a month to $10,000 a month, depending mostly on how well known and how successful they are. Agencies selling to executives sit at the top of that range. (For the full breakdown, I wrote a separate post on LinkedIn ghostwriter cost, and there's a ghostwriter cost calculator if you want to run your own numbers.)
Here is what the typical agency process looks like:
- They interview you for about 30 minutes, every two weeks.
- A writer who has usually never worked in your industry turns that into posts "as you."
- An account manager sends drafts. You comment. They revise. Repeat for every post.
- Approved posts go out on a schedule.
Look at where the hours go. Almost all of the labor sits in steps 2 and 3: drafting and the back-and-forth on each draft. That is exactly the work AI now does well.
And the input in step 1 is thin. Thirty minutes every two weeks is a small window into what a founder actually thinks. The most common complaint I heard on calls was some version of "my ghostwriter sounds nothing like me." The second most common came from companies with several executives: "we'd need five agencies for five executives." Neither problem gets solved by hiring a better writer.
Why AI ghostwriting wins on the drafting layer
An AI ghostwriter is better at drafting for three structural reasons.
It can listen all the time. A human writer gets your voice through a biweekly interview. An AI system can take in your voice memos, call notes, old posts, and the news you react to, every day. More signal in means posts that sound more like you. That's the difference between interview-based ghostwriting and continuous capture, and it's why I built voice memos into my own workflow.
It has no queue. A ghostwriting agency runs revisions through an account manager, so you wait for their availability. With an AI ghostwriter you comment on a draft, get the next version in seconds, approve it the same day, and it's sealed and scheduled. Nothing goes out until you approve it.
It learns across customers. This is the part I find most interesting as a builder. A single ghostwriter can personally handle maybe ten clients a year. They can't see patterns across hundreds of founders. A product serving a few hundred customers can see what worked for each segment, which audiences respond to which kinds of posts, and feed that back into every new draft. No individual writer can match that, however talented.
Voice matching on its own is not enough, by the way. We once matched a CEO's voice almost perfectly and he told us it sounded like every CEO on LinkedIn. The drafting layer only works when the substance comes from you. AI is great at the writing. It still needs your raw thinking to write about.
What humans keep: accountability, strategy and trust
Here is where I think most "AI will kill ghostwriting" takes get it wrong.
I had a long conversation with the owner of a small personal-branding agency. He ghostwrites two posts a week for founders at $1,500 a month, has about 20 customers, and has been running for a bit over a year with pretty stable retention. I asked him what his clients actually value most.
His answer was accountability. He compared himself to a personal trainer: without the trainer, you would not go to the gym twice a week. His clients would not post twice a week without him.
That stuck with me, because it means a big part of what a ghostwriter sells has nothing to do with writing. So here is how I'd split the job:
| Part of the job | Who does it better now |
|---|---|
| Drafting posts | AI |
| Revisions and back-and-forth | AI |
| Capturing your ideas continuously | AI |
| Scheduling and consistency tracking | AI |
| Pattern data across many clients | AI (at scale) |
| Setting the narrative and positioning | Human, with AI support |
| Accountability that you actually feel | Human |
| Pushing back on a bad take | Human, for now |
| Knowing your industry from the inside | Depends on the human |
The human column is shorter, and it's the column founders pay a premium for.
The pricing trap: why founders pay a person and refuse the software
There's a strange piece of buyer psychology that decides a lot of this market, and I learned it the hard way.
A founder told me he had just hired a person to handle his LinkedIn posting and commenting. I asked him if he would pay $400 a month for software that did the same job. He said I was crazy, that's way too much for a tool. I pointed out he had just hired a person to do exactly that. His answer: paying $400 for a person is different, because the person evolves and the software doesn't.
That's mental accounting at work. When you hire a ghostwriter, you see the effort. Someone sits with you for a four-hour strategy workshop. Someone sends you drafts, reads your comments, and comes back. You can watch person-hours being spent on you, and that makes a monthly retainer feel fair. When the output is "posts from software," the founder compares it to ChatGPT at $20 a month and the price feels absurd, even when the posts are better.
The uncomfortable corollary: some of that friction is what makes the service feel valuable. Remove the back-and-forth, and a client may ask why they're paying so much for eight posts a month. Ghostwriters who survive the AI shift will be the ones who can point at value beyond the drafts.
Will AI replace ghostwriters? The model that replaces the agency
My honest prediction: the ghostwriting agency built on biweekly interviews and per-post revisions will shrink. What replaces it is a hybrid.
- A human sets the strategy. A real session up front to define the narrative, the angle and the content plan, then periodic check-ins, maybe quarterly.
- An AI ghostwriter does the daily work: captures your ideas, drafts in your voice, revises on your comments, schedules what you approve.
- You stay the approval gate. Nothing ships without you.
That's roughly the model I landed on after going back and forth on it for months. The strategy conversation is the part I refuse to hand to software, because that's where someone who has done this before earns their keep.
There's a second version of this that I think is underrated: the ghostwriter who uses AI. A one-person agency supporting ten accounts that cuts drafting time in half can take on twice the clients without hiring. For a good ghostwriter, AI speeds up the part of the job that was always the bottleneck. The best-case future for a talented ghostwriter is a founder paying them for strategy and accountability, with both of them working inside the same AI system.
Should you hire a ghostwriter or use an AI ghostwriter?
Here's how I'd decide, based on the calls I've been on:
- Use an AI ghostwriter if you have things to say, you'll show up to approve drafts, and your main problem is time. You'll get more posts that sound like you, for less.
- Hire a human if you honestly will not post without someone chasing you, or if you need real positioning work and a thinking partner more than you need drafts.
- Do both if you run content for several executives. One strategist plus an AI ghostwriting system beats five separate agencies for five executives.
Whatever you pick, check the output. Paste a draft into an AI-sounding post checker and see whether it reads like you or like the median LinkedIn post. That's true for AI drafts and for human ones.
The thing I keep coming back to is my own rule: you can't lead what you've never touched. Founders who hand the whole thing to anyone, human or machine, and never engage with it end up with a feed that isn't theirs. The ones who stay hands-on, feed it their real thinking, and approve every post get the benefit of AI without losing their voice.