Is It AI Slop…or just bad content?

LinkedIn has a new button. Open the menu on a post and, alongside the usual reporting options, you may now see: “Seems like AI slop.”

Yes, LinkedIn is officially letting users call out content they think is AI-generated and low quality.

The feature is part of a wider effort to clean up the LinkedIn feed. The platform says it wants to reduce generic posts, automated comments and content that looks polished but adds little original perspective. (LinkedIn News)

And judging by the reactions, plenty of people are ready for it.

Some want the same button on YouTube. Others are celebrating the possibility of seeing fewer generic AI posts and comments.

But there is one pretty big problem: What exactly counts as AI slop?

Because the more you think about it, the harder that question becomes.

‍

What does “AI slop” actually mean?

Before we start reporting AI slop, we probably need to agree on what it is. LinkedIn describes it as low-effort, AI-generated content that may look polished but lacks unique perspective or substance. It also makes an important distinction: using AI to help you write is completely fine. The value should simply come from your own voice and perspective.

So AI-generated content isn't automatically AI slop.

Which leaves us with a more interesting question:

What actually makes something slop?

Is it the fact that AI was involved? The lack of original thought? The low effort? The lack of useful information? Because those things aren't interchangeable.

Someone could spend 30 seconds prompting AI and publish a genuinely useful answer. Someone else could spend two hours manually writing a post that says absolutely nothing.

And that's where the definition starts to get messy.

If low quality is what makes content slop, then AI isn't really the defining characteristic at all.

And if AI-generated is what makes it slop, we're judging the process rather than the result.

Maybe the problem isn't AI content. Maybe it's just bad content.

‍

Maybe we're asking the wrong question

The interesting part of LinkedIn's definition isn't actually the word AI. It's everything else:

Low effort. Generic. Repetitive. No unique perspective. No substance.

None of those things were invented by AI.

LinkedIn had generic motivational posts long before ChatGPT.

We had recycled business advice. Engagement bait. Empty thought leadership. People turning one relatively simple observation into 17 paragraphs. Posts beginning with a dramatic personal story before somehow ending in a sales pitch.

All written by humans.

One reaction to LinkedIn's new feature summed up the problem rather nicely:

🤣

Which raises an uncomfortable question.

If someone spends an hour manually writing a generic post that adds absolutely nothing new, is that automatically better than a useful post someone created with the help of AI?

Probably not.

‍

AI didn't invent bad content. It made bad content scalable.

This is where AI really has changed the equation. Before generative AI, producing mediocre content still required effort. You had to sit down. Write the post. Write another post tomorrow. Think of something to comment. Write the comment. Repeat.

AI can dramatically reduce that effort.

One generic idea can become 20 posts. One automated system can produce hundreds of comments. One recycled opinion can be rewritten over and over again.

LinkedIn itself points specifically to AI being overused at scale and in automated ways as a problem. It's also targeting comments created at scale with little or no human involvement and replies that simply restate the original post without adding anything. That's an important distinction.

The problem isn't simply: AI → bad content

It's closer to: Low effort × generic ideas × AI × scale = a feed full of slop

And that is a genuine problem.

When creating something costs almost nothing, there is very little friction stopping people from creating a lot of nothing.

‍

But AI can also make good ideas easier to share

There's another side to this. Not everyone with something valuable to say is a great writer.

A founder might have 20 years of experience but struggle to turn those experiences into a concise LinkedIn post. Someone writing in their second or third language might have a genuinely original perspective but use AI to communicate it more clearly. A marketer might conduct the research, develop the argument and provide the examples, then use AI to organize everything.

AI can lower the barrier between having a good idea and communicating that idea well.

That's very different from asking AI to manufacture an opinion for you.

And LinkedIn appears to recognize that distinction. Its own guidance isn't “don't use AI.” It says AI can help with writing, while the voice and perspective should still come from the person behind it. So perhaps the better test isn't:

Did AI write this? It's: Did the person have something to say?

‍

Then we have another problem: who decides what “low quality” means?

Let's say we agree that AI slop means low-quality AI content.

Great.

Now define low quality.

A basic post about five ways to improve your LinkedIn profile might feel painfully obvious to someone who has worked in social media for ten years. To someone creating their first LinkedIn profile, it could be genuinely useful. A post could be beautifully written and say nothing new. Another could be badly written but contain an insight you've never considered before. One person might see an AI cliché. Another might simply see someone's natural writing style.

And now millions of users potentially get to make that judgment themselves.

LinkedIn says its systems have been trained to recognize signals of generic content and look for perspective, context and expertise. The company reported that in its initial testing, it was correctly identifying generic content 94% of the time. But the reporting button introduces human judgment into the equation too.

And humans aren't exactly objective classifiers.

‍

What happens when people start reporting content they simply don't like?

This is where things could get particularly interesting.

Imagine you see a competitor's post performing extremely well.

“Seems like AI slop.”

Someone posts an opinion you strongly disagree with.

“Seems like AI slop.”

A creator has a very polished writing style that feels AI-ish to you.

“Seems like AI slop.”

Could people eventually automate the reporting itself?

Could someone tell an AI agent to go through competitors' posts and flag them?

That's speculative, of course. But once subjective user reporting becomes a signal, abuse is an obvious question platforms have to account for. LinkedIn says the reports will help it tune its models and improve what appears in people's feeds. (Business Insider) That means the quality of the human signal matters. Because we're essentially asking humans to help AI get better at identifying bad AI. There's something wonderfully 2026 about that.

‍

Maybe we need a “human slop” button too

Perhaps we're overcomplicating this. Instead of:

“Seems like AI slop”

LinkedIn could add a second option:

“Seems like human slop.”

Possible reasons for reporting:

• Generic motivational story you've read 37 times
• Eight paragraphs without a discernible point
• “Here are 5 things entrepreneurship taught me”
• Completely recycled opinion
• Fake vulnerability followed by product pitch
• Comment that just says “Great insights!”
• Post ending with “Agree?”

Suddenly the distinction becomes a little ridiculous.

Because if the goal is a better LinkedIn feed, does it really matter whether bad content was produced by a human or a machine?

There is a reason the conversation around AI slop matters. Generative AI makes it possible to flood platforms with generic content at a scale we haven't had before. Automated comments and mass-produced posts can absolutely make social platforms worse.

But AI is also becoming part of normal creative workflows.

It can help research a topic, challenge an argument, structure someone's thoughts, improve writing, adapt an idea for different platforms or turn expertise into content faster.

At Whaaat AI, for example, our specialized agents are designed around different marketing tasks and work with brand context rather than treating every request as an isolated generic prompt. The goal isn't to remove the human from marketing. It's to remove the repetitive work around it.

The technology isn't what determines whether the final result is worth reading. The thinking does. Maybe that's the distinction we should actually care about. A human can produce slop. An AI can produce slop. And a human using AI can produce something genuinely useful.

So yes, give us tools to reduce generic, automated content. But perhaps we should be careful about treating AI-generated and low-quality as though they mean the same thing.

Because they don't.