Is Thought Leadership Dead? The AI Slop Machine.

Scroll your LinkedIn feed for ninety seconds and count the ‘thought leadership’ posts that have a catchy one-line hook, some bullet points, and close with ‘Agree?’ or some equally glib call to action. You can’t always be sure which ones were written by AI, but now you ignore most of them anyway, having stopped believing a while ago that a human sat down, thought them through, and wrote them. Especially given how often they’re pumped out in some cases. So the only real question is how bad the AI slop really is.

Well, it is worse than you imagined. Pangram, an AI-detection firm, found that more than 40% of long-form LinkedIn posts are now flagged as entirely AI-generated. Stop and pause on that. Not partially, but entirely AI-generated. And even more interesting is that, of all posts flagged as AI-generated across platforms, 62% came from LinkedIn. This means AI-generated content now saturates the feeds where professional discourse and thought leadership are supposed to be happening.

Let’s be blunt about this. It is a brilliant example of AI being used badly and lazily, at scale, by people who should know better. Or maybe they don’t. Or maybe they just don’t care in their pursuit of credibility, engagement, and leads.

A 2026 Use.AI survey of 7,200 people found that 69% of people felt it was ok to publish AI-written work under their own name if they gave the initial prompt, and that 46% would publish AI-written work under their own name without disclosing it. 54% felt any AI involvement should be disclosed. It appears we demand authenticity from the people we read, and quietly reserve the right to fudge our own.

And interestingly, it is not just individuals who are doing this. Earlier this year, the Consulting Big Four were accused of publishing “thought leadership” that was allegedly largely AI-generated. This is in an industry that has built its entire business model on the promise of hard-won real-world expertise. These companies are employed to deliver insights, yet they allegedly outsourced it.

In March 2026, Harvard Business School fellow John Winsor argued in the Harvard Business Review article “Has AI Ended Thought Leadership?” that the category itself is dying. That AI has made it incredibly easy to perform expertise without possessing it, saturating the feed with polished, confident thought leadership content that is not based on real expertise. And that is the problem. AI writes fluently and confidently, and it projects expertise, but when you want to engage that expertise based on those thought leadership posts you have read, it often isn’t actually there.

Genuine thought leadership and expertise emerge from the scar tissue of being in the trenches. The failed product launch, the brutal downsizing you had to lead, the moment the negotiation turned, the comeback from the edge, and the constant adaptation and pressure to perform. AI has none of this. What it has comes from its LLM synthesis of available content. But it lets people promote “expertise”, confidently, with nothing to back it up.

Even LinkedIn has recognized how damaging this is. In May 2026, it downranked AI-generated content and removed its AI writing assistant from the post button, with editorial VP Laura Lorenzetti admitting that overused AI “dilutes the valuable insights that real human conversations can spark”. By July, it launched a “Seems like AI slop” reporting button.

So here’s the truth of it. Don’t outsource your thought-leadership to AI. You know it when you see it, and so do other people. Take the time. Use AI to sharpen your sentences and grammar, but not to supply the thought. Help stop the AI slop.