LinkedIn Is Cracking Down on AI Slop. This Is Good News for Employee Advocacy

LinkedIn is reducing the visibility of AI slop

In a recent article, LinkedIn VP of Product Laura Lorenzetti explained that LinkedIn is taking steps against low-effort AI-generated content that looks polished, but lacks real perspective or substanceLinkedInLinkedIn: Keeping conversations real on LinkedIn — Laura Lorenzetti (2026)Opens in a new tab.

The guidance is now written down, not just an opinion piece

LinkedIn’s Help Centre has since published best practices for content created with the help of AILinkedInLinkedIn Help Centre: Best practices for content created with the help of AIOpens in a new tab. It defines AI slop, asks that every AI-assisted post be reviewed and approved by the person publishing it, and gives readers a way to flag content that reads like slop.

AI itself is not the problem, fake automation is

As Lorenzetti points out, AI can still help people write. The issue starts when AI is overused at scale, in an automated way, with little human involvement. That includes automation tools, fake AI profiles and comments that simply restate the original post without adding anything new.

Employee advocacy needs real employee context

Marketing can still guide the message, campaign and strategic themes, but the value comes from real employees adding their own tone, role context, experience and judgement.

The opportunity is scalable human contribution

The answer is a better process, one that helps real employees share what they know, see and believe with less friction. More copy-paste posts and fully automated content take companies further from that.

For a while now, LinkedIn has had a bit of a content problem. Not because more people are using AI, but because more content is being published without much real input from the person behind the post.

You probably feel it in your own feed. A post looks polished, structured and professional at first glance. The formatting is clean. The sentences are smooth. The message is easy to follow. But after a few lines, something is missing: no real context, no specific experience, no clear opinion, no sign that the person has actually seen, learned or struggled with the thing they are writing about.

That is what many people now call AI slop. Low-effort content that may look good on the surface, but lacks real perspective or substance. Slop is what happens when AI is used to produce content at scale without enough human input behind it, which is a different thing from using AI to write better.

This is now becoming a platform issue too. In her article “Keeping conversations real on LinkedIn”LinkedInLinkedIn: Keeping conversations real on LinkedIn — Laura Lorenzetti (2026)Opens in a new tab, LinkedIn VP of Product Laura Lorenzetti writes that LinkedIn is taking steps to reduce low-value, generic content and comments. She also points to automation tools, fake AI profiles and low-value comments as part of the same broader problem.

For B2B marketing teams, that signal carries weight. LinkedIn is where buyers form opinions about people, companies, categories and expertise. If generic content gets less space, then the way companies activate employee advocacy has to change too.

One detail is worth noting here. LinkedIn is not banning AI content, and the platform is clear that people can use AI to help them write. What it asks for is content that carries real perspective, context or expertise.

Laura Lorenzetti puts it clearly: “It’s ok to use AI to help you write, but your posts and comments need to represent your voice and your perspectives.” That is the distinction companies need to understand. The problem starts when the human is removed from the process.

LinkedIn also says it is training systems to recognise signals of AI slop and identify content that feels generic or repetitive, even if it looks polished on the surface. In its initial testing, LinkedIn says it is correctly identifying generic content 94% of the time. That should get the attention of anyone who relies on LinkedIn for visibility.

This also goes beyond posts. LinkedIn is looking at comments created at scale, responses that simply restate the original post and fake AI profiles that imitate human engagement without bringing any real experience, relationship or point of view. That is a different problem from an employee using AI to improve a post. One supports a real person. The other fakes one.

A post can still be written with AI support and be useful. A person may use AI to organise their thoughts, improve wording or get past the blank page. For many professionals, especially people who are not native English speakers, that can be a very helpful bridge.

But the final post still has to sound like it came from a person. It needs a point of view. It needs context. It needs some trace of actual experience.

That is the real shift.

Since that article, the position has moved from an opinion piece into published guidance. LinkedIn’s Help Centre now carries a page setting out best practices for content created with the help of AILinkedInLinkedIn Help Centre: Best practices for content created with the help of AIOpens in a new tab, and it repays a close reading, because it describes the working method LinkedIn expects rather than the technology it objects to.

The same page defines what LinkedIn is acting against: “low-effort, likely AI-generated content that may sound polished on the surface but lacks a clear point of view, unique perspective, or substance”. It is worth noticing what that definition leaves out. Nothing about which tool produced the draft. Nothing about whether AI was involved at all. The test is the perspective inside the post, not the software behind it.

The best practices read like a description of a person in the loop

• Follow the Professional Community Policies, the same rules that apply to any other post.
• Review, edit and approve AI-assisted content before it goes out. You stay responsible for everything you publish on LinkedIn.
• Respect intellectual property and privacy rights when you use AI tools.
• Be transparent about heavy reliance on AI when that is not obvious from the context.

Every one of those items assumes a named person who saw the draft, changed what was wrong and decided to publish it. None of them can be satisfied by a system that writes and posts on somebody’s behalf while they are sitting in a meeting.

LinkedIn also states the consequence plainly. Content that is generic, repetitive or lacking a clear point of view is, in its own words, less likely to be widely distributed. Readers now have a hand in that as well: a “Seems like AI slop” option in the post menu lets members flag what they see, and authors can find a “Seems like AI” note in their post analytics. The post is not taken down. The feedback simply becomes visible to the person whose name is on it.

For employee advocacy, the word to underline there is repetitive. One campaign text, sent out unchanged and published by a row of colleagues in the same week, is close to a textbook example of it. That is also the model a great many advocacy programmes still run on.

Which makes this guidance far easier to work with than the headlines suggest. LinkedIn is not asking companies to stop using AI, and it is not asking employees to write everything from scratch. It asks that a real person’s perspective ends up in the post, and that a real person reviews it before it goes out. That is a workflow question, and workflows can be fixed.

There is a big difference between AI support and AI slop. AI support starts with something real: a customer conversation, a product decision, a lesson from a project, a founder’s opinion, a sales objection that keeps coming back. AI can then help structure that idea and make it easier to publish.

AI slop starts when the output becomes the whole process. A prompt goes in. A polished post comes out. Nobody adds context. Nobody challenges the wording. Nobody asks whether the person whose name is on the post would actually say it that way.

The same risk exists in employee advocacy. A company may have something valuable to promote: a report, an event, a customer case, a product update. That campaign message is worth sharing, but on its own it is usually not enough for a strong personal post.

The difference comes from what is added around it. Why does this topic matter to this employee? What do they see in their role? What customer conversations, product decisions, hiring challenges or market signals make the campaign relevant to them?

Without that layer, even a well-written post can feel interchangeable. You can often remove the author’s name and company, and not much changes. The post still works as a piece of professional content, but it does not tell you enough about the person behind it.

LinkedIn is valuable when people share what they actually see in their work. Those posts do not need to be perfect. They need to be specific.

For years, many employee advocacy programmes have followed a very basic model. Marketing creates a campaign. The campaign needs reach. Employees receive a suggested post and are asked to share it.

Sometimes everyone receives the exact same text. Sometimes there are a few small variations. But the underlying idea is still the same: marketing writes, employees distribute. That approach was already limited.

It often creates posts that feel too close to the company page. Employees hesitate because the content does not sound like them. Their networks can also see what is happening, especially when several colleagues post almost the same thing on the same day.

Now there is another reason to rethink it. If LinkedIn is becoming more careful with AI slop, automated comments and low-value content, copy-paste advocacy becomes even weaker. It does not create strong conversations, and it does not help employees build a credible presence.

That does not mean companies should stop running campaigns. It means campaigns need to become more personal. Marketing can still define the theme, message, audience and goal. That structure is useful. Without it, employee advocacy becomes scattered very quickly.

But employees need more than a campaign message. A salesperson should be able to connect the topic to customer conversations. A recruiter should be able to connect it to the candidate experience. A product manager should be able to explain what it means from a product or market perspective. A founder or executive should be able to connect it to the company’s direction.

For marketing teams, this is the bigger opportunity. Employee advocacy should not be treated as a cheap distribution trick. It is a way to make the company visible through the people who actually carry its expertise, customer knowledge and point of view. This is the same reason internal experts are usually a stronger source of thought leadership than the brand voice itself.

That does not happen automatically. It requires input, context and editing.

This is where the conversation often gets practical very quickly. Most companies do not want copy-paste advocacy, but they also do not have the time to manually interview every employee, write every post from scratch, review every sentence and coordinate the entire calendar by hand.

So the answer is usually somewhere in the middle. Marketing gives direction. AI or a writer helps create a first version. The employee adds context, edits what feels off and decides what belongs on their profile. Not perfect. But much better than pretending that either full automation or fully manual writing is the only serious option. Different roles will land on different angles, which is also the heart of activating sales, HR and product teams from one company strategy.

A practical worry usually follows: if a tool is involved in publishing, does the post get seen by fewer people? Asked that directly, LinkedIn’s own feed product lead answered it.

“Does posting from a 3rd party tool impact my reach? No. It’s all about matching the right content to the right audience.”

View the post on LinkedInopens in new tab

RJ

Rishi Jobanputra

Senior Director of Product Management, Feed at LinkedIn

“Does posting from a 3rd party tool impact my reach?” No. It’s all about matching the right content to the right audience. Let’s test this out from CapCut.

see more

16431 comments · 13 reposts

That distinction shapes the whole discussion. What LinkedIn is acting on is generic, low-value content, not which tool was used to help draft the post.

This is one of the reasons we built Heyoo the way we did. Employee advocacy should not mean sending one post to everyone and hoping people share it. That approach may create activity, but it does not create much credibility.

Heyoo is built around real employees. There is no fake activity, no fake profiles and no automated engagement. Marketing creates a campaign around a topic, asset, event or strategic theme through Advocacy Campaigns. Employees then receive personal post suggestions based on their role, personal tone of voice and campaign context.

That does not mean the campaign message disappears. It means the campaign is not treated as the whole story.

Heyoo does not make a post authentic by itself. No tool does. What it can do is give employees a better starting point than a blank page or a copy-paste campaign post.

From there, the employee stays in control. They can edit the post, adjust the wording, add their own experience, remove what does not feel right and decide what belongs on their profile.

Held against LinkedIn’s own checklist, that sequence lines up item by item. The draft starts from the employee’s role, experience and voice, so the perspective in the post is theirs. Nothing reaches the feed until they have reviewed and approved it, which is what the approval workflow is built for. And because colleagues each start from their own angle on a campaign instead of from the same paragraph, the posts differ from one another by design, which is the opposite of the repetition LinkedIn says it distributes less widely.

For leaders and subject matter experts, marketing can provide more active support through Managed Profiles. That helps busy people stay visible without having to start from scratch every time. But here too, control and approval remain important. Thought leadership works best when the person behind the profile still owns the message.

So no, the future is not more generic AI content. And it is also not a return to everyone writing everything manually from zero. The better path is somewhere in between: marketing provides the direction, AI reduces the friction and employees add the perspective.

For marketers, that is the important shift. Advocacy becomes less about asking people to distribute content and more about helping them participate in the market conversation with something that actually sounds like them.

Before you scale a programme, it helps to know what it is worth. The employee advocacy ROI calculator models earned media, pipeline and the time your team saves.

LinkedIn’s move against AI slop should worry anyone who has taken the human out of the process. For companies using AI to help real employees write, it is good news.

For B2B companies, that is a healthy push. Employee advocacy was never meant to be fake activity, automated engagement or copy-paste distribution. It should help real people share what they know, see and believe.

That is harder than generating more content.

But it is also much more valuable.