How to use AI for LinkedIn content without losing your voice

If you've spent any time on LinkedIn recently, you've probably come across posts that you can tell have been written almost entirely by AI.

There's nothing necessarily wrong with using AI to help create content. I use it within my own business and it can be brilliant for brainstorming, organising ideas or helping you get started when you've got lots of thoughts but no idea how to turn them into a post.

The problem comes when we stop at the first draft.

LinkedIn has recently introduced a new reporting option for posts that appear to be heavily AI generated or low quality, which I think tells us something interesting about where content is heading.

This is about making sure we're still adding something of ourselves before we hit publish.

What has LinkedIn actually changed?

LinkedIn users can now select "Seems like AI slop" when reporting a post. LinkedIn has also been working on how it identifies low quality content and what gets recommended to people outside your immediate network.

Interestingly, LinkedIn is also replacing an AI feature that rewrote posts with a simpler proofreading tool designed to keep more of the writer's original voice, which I think is the more interesting part of this update.

We've spent the last few years talking about everything AI can do for us. Now we're starting to see more conversation around how we use it without making everything online sound exactly the same.

Why does this matter for small businesses?

If you're using LinkedIn to market your business, you're not simply trying to prove that you know your stuff. You're giving potential clients an idea of who you are, how you work and whether you're someone they could see themselves working with.

That's particularly important for service based businesses where people are often buying into the person as much as the service.

You could publish a perfectly written post about "unlocking your potential" and it might technically tick all the boxes. But if another 500 businesses could have written it, what is someone actually going to remember about it?

I'd much rather read about the question three clients asked you this week, something you changed your mind about recently or a lesson you've learnt from actually doing the work.

Those are the details that make your content yours.

How to use AI for LinkedIn without sounding generic

I don't think AI is the problem. Using AI doesn't automatically make your content generic. You can use it to brainstorm ideas, organise a rambling voice note, help structure a blog, repurpose something you've already written or give you somewhere to start when you're staring at a blank page.

All of those things can save a huge amount of time.

But I wouldn't give AI a vague prompt, copy whatever comes back and publish it without touching it. That's where you start losing the phrases, opinions and little details that make you sound like an actual person.

Think of the first AI generated version as a draft rather than the finished piece, then go back through it and make it yours.

iPad and stylus on a café table used for planning LinkedIn and social media content

A really simple test I use

One of the easiest ways to spot AI-ish writing is to read your content aloud.

Would you actually say that sentence to someone?

I notice this particularly with phrases that sound impressive written down but aren't something you'd ever naturally say in conversation.

If you're reading your LinkedIn post thinking, "I have literally never used that phrase in my life", then change it.

You can also add something AI wouldn't automatically know about you. A conversation you've had with a client, something you've noticed in your industry, an opinion you have or an example from your own business can completely change how a post feels.

Grace from Silver Lining Social working on her laptop and planning digital marketing content

Before you publish your next LinkedIn post

If you've used AI to help write it, I'd do a quick final check:

  • Read the whole thing aloud and change anything that doesn't sound natural.

  • Look for vague or overly polished phrases and simplify them.

  • Add at least one genuine example, experience or opinion.

  • Check that the opening actually tells people what the post is about.

  • Fact check anything AI has introduced that you didn't provide yourself.

  • Ask yourself whether someone who knows you would recognise your voice.

You don't need to deliberately make your writing imperfect to prove a human wrote it. You just need to make sure you've actually contributed something to it.

Use AI for the bits it's good at

I don't think small business owners need another thing to feel guilty about and using AI to make your marketing workload easier isn't something I'd tell you to stop doing.

Use it where it genuinely saves you time.

Let it help you get past the blank page, organise your ideas or turn messy notes into something you can work with, but keep your judgement, experience and personality involved in the final version.

Ironically, the more AI generated content we see online, the more valuable a recognisable human voice becomes.

The Verge covered LinkedIn's new reporting option on 30th July 2026.

If you're using AI for your marketing but aren't quite sure how to turn all those ideas into an actual strategy, a Social Brew can help you work out what deserves your attention and what can wait.





Next
Next

What caught my eye in July: social media and marketing updates