AI can draft LinkedIn posts that sound like a founder, but only when it works from that founder's real material: their past posts, recorded talks, opinions and stories, gathered into a voice file. The AI drafts from the file and the founder edits and approves every post. Nothing goes out under their name without them.
Key takeaways
- A founder's voice comes from their own material, not from a generic "professional but friendly" prompt.
- A voice file records how they write, what they believe, the stories they tell and the phrases they avoid.
- According to Edelman and LinkedIn, about 73% of B2B decision-makers trust thought leadership more than marketing materials for judging a company.
- The founder edits and approves every post; AI never posts on their behalf.
- According to LinkedIn's help pages, bots and browser extensions that automate activity on the site are not permitted.
Can AI write LinkedIn posts that sound like you?
Yes, if it has enough of your real material to learn from. Given a dozen of your best posts, a few recorded talks and your positions on the topics you care about, AI can draft posts that use your sentence rhythm, your examples and your opinions. Without that material, it produces the generic LinkedIn voice everyone recognises.
The generic voice is easy to spot: a one-line hook, short dramatic paragraphs, a lesson in a list and a question at the end. Readers scroll past it because they have seen it a thousand times. Macquarie Dictionary named "AI slop" its 2025 Word of the Year, and LinkedIn feeds are where many people meet it.
A founder's real voice is different in small ways that add up. They might use long sentences, swear occasionally, always start with a customer story or never use lists. Those habits are what the AI needs to learn.
What is a voice file?
A voice file is a short document describing how one named person writes and thinks. It records their tone, sentence habits, recurring stories and analogies, firm opinions, phrases they use and phrases they would never use. It is built from their own posts, talks and interviews, and only they can confirm it is right.
| Section | What it holds |
|---|---|
| Register | Formal or casual, short or long sentences, humour |
| Positions | Opinions they hold firmly, and ones they avoid |
| Stories | The anecdotes and examples they return to |
| Phrases | Words they use often, and words they never use |
| Limits | Topics they will not post about |
It is separate from the company's brand voice. A company page speaks for the organisation; a founder's posts speak for one person. The guide to brand voice guides for AI covers the company side.
How should AI draft founder posts?
AI should draft from three inputs: the voice file, a specific idea worth posting, and evidence such as a customer conversation, a result or a lesson. It writes a draft in the founder's voice, checks it against the voice file's rules, and puts it in a review queue. The founder then edits and approves.
A useful weekly rhythm has three steps. First, gather raw ideas from the founder's week, such as calls, meetings and things they said in passing. Second, turn the strongest into a handful of draft posts. Third, give the founder a short review session to edit, reject or approve.
The idea matters more than the writing. A post built on a real customer conversation or a hard decision beats a polished post about nothing. The AI's job is to turn the founder's raw thinking into a draft quickly, not to invent opinions for them.
What should a founder always do themselves?
A founder should approve every post, reply to comments in their own words and decide which opinions to publish. They should also handle messages from prospects and anything sensitive, such as hiring, customers or competitors. AI can prepare drafts and suggestions, but the founder's name means the founder decides.
Posting itself should stay manual. According to LinkedIn's help pages, members may not "use bots or other unauthorized automated methods to access the Services", and accounts that do "risk having their accounts restricted or shut down". Drafting in a separate tool and posting by hand avoids that risk entirely.
Direct messages need the most care, since they come from a real person's profile to another real person. The guide to LinkedIn outreach risks covers why.
How do you keep AI-drafted posts authentic?
Keep them grounded in things that really happened, in opinions the founder really holds and in words they would really use. Let the founder rewrite freely, and feed their edits back into the voice file so drafts improve. If a post would embarrass them when a customer asks about it, it should not go out.
Trust is the reason this matters. According to Edelman and LinkedIn's 2024 B2B Thought Leadership Impact Report, about 73% of decision-makers say thought leadership is a more trustworthy basis for judging a company's capabilities than its marketing materials. That trust disappears quickly when posts feel hollow.
The edits are the best training data. Every phrase a founder deletes and every sentence they rewrite tells the system something. The guide to AI for founders covers using the same voice file beyond LinkedIn.
Skein builds a voice file for each person who posts under their own name and drafts from it, with that person approving every post before it goes out; the AI engineering page shows the LinkedIn workflow. That suits a founder with opinions and too little time to write. A founder who enjoys writing every post may only want an editor.
Frequently asked questions
Can AI write LinkedIn posts in my voice?
Yes, if it learns from enough of your real material, such as past posts, recorded talks and your stated opinions. Gathered into a voice file, that material lets AI draft posts with your rhythm, stories and views. You should still edit and approve every post before it goes out.
What is a founder voice file?
A founder voice file describes how one person writes and thinks: their tone, sentence habits, recurring stories, firm opinions, favourite phrases and phrases they avoid. It is built from their own posts, talks and interviews, kept separate from the company's brand voice, and confirmed by the founder.
Is it okay to use AI for LinkedIn posts?
Using AI to draft posts that you then edit and approve is widely accepted, especially when the ideas and opinions are genuinely yours. Avoid tools that post or engage automatically, because LinkedIn prohibits bots and automation on its site, and accounts using them risk restriction.
How do I stop AI LinkedIn posts sounding generic?
Give the AI your own material rather than a style instruction, base each post on something that really happened, and ban the patterns you dislike, such as one-line hooks or forced questions. Feed your edits back into the voice file so future drafts sound more like you.
Should AI reply to LinkedIn comments for me?
No. Comments and messages are conversations with real people who expect to hear from you. AI can suggest points to make, but replies should be written or approved by you in your own words. Automated engagement also breaches LinkedIn's rules on bots and automation.
Related reading
- Edelman and LinkedIn. "2024 B2B Thought Leadership Impact Report." Edelman, 2024. https://www.edelman.com/insights/thought-leadership-gets-b2b-buyers-back-into-game
- LinkedIn Help. "Prohibited software and extensions." LinkedIn, accessed 30 October 2026. https://www.linkedin.com/help/linkedin/answer/a1341387
- The New Daily. "'AI slop' is Macquarie Dictionary's word of the year." The New Daily, 25 November 2025. https://www.thenewdaily.com.au/life/2025/11/25/ai-slop-is-word-of-the-year