On July 30, LinkedIn rolled out a button that lets you flag a post for "seeming like AI slop." It got over a million clicks in the first two weeks. That tells you something: we’ve gotten so anxious about AI that being mistaken for a robot has become the insult.

I use AI every day. I train other teams how to implement it in their businesses. And when I’m confronted by the “AI slop” button, or a negative comment about AI, I find myself minimizing my own use of it. So I wanted to know if that instinct was rational, or just absorbed from the noise.

The data says it’s rational. A 2025 Duke University study found that colleagues rate AI users as lazier and less competent than people completing identical work manually. A 2026 Atlassian study went further: people who disclosed their AI use were rated ten times lazier than people who didn’t say anything at all, and were 24 percentage points less likely to get recommended for a high-visibility project. Depending on the survey, somewhere between a third and well over half of workers say they hide their AI use from their employer, and the number climbs, not falls, the higher up you go. In one U.S. study, over half of C-suite leaders admitted to concealing it. The question isn’t whether the shame is real. It’s where it comes from, because most people can’t name it beyond a general unease.

Here’s the piece I think gets missed. We still culturally reward the appearance of effort. A quote that took four hours to write reads as more legitimate than one that took twelve minutes, even when the twelve-minute version is better, because we’ve spent a hundred years equating visible struggle with competence. AI breaks that signal. It doesn’t remove the hard part of the work; it removes the part that used to make the hard part visible. What’s left, the judgment about what’s worth saying and whether the output is right, is invisible unless you show it on purpose. As a result, people don’t disclose the tool, because disclosing the tool without disclosing the judgment reads as “I didn’t do the work,” even when the judgment is the entire job.

The second layer to this conversation is the jobs anxiety, and it deserves to be taken seriously. Entry-level roles are getting squeezed; Stanford’s 2026 AI Index found employment for software developers aged 22 to 25 down nearly 20% since 2024. If that’s your mental model of what AI means, then telling your manager you used it daily doesn’t sound like a good idea. But the same data that shows displacement also shows re-composition happening in parallel: LinkedIn logged 1.3 million new AI-related roles globally in two years, among them AI engineers, forward-deployed engineers, and data annotators, roles that barely existed five years ago. PwC found productivity growth in AI-heavy industries has nearly quadrupled since 2022.

The skill that’s rising in value isn’t "can you do the task." It’s "can you direct the AI and know whether what it gave you is actually good." It’s an entirely new job, extending existing skill sets at a higher altitude, and almost nobody frames the disclosure conversation that way. The experts in their field aren’t at risk of losing their jobs because AI will take them. The AI needs their expertise to be able to function. What AI does, in reality, is it transforms jobs. It takes the repetitive action off the human, so the human can elevate their thinking, guide the AI tools, and envision what else is possible when they aren’t up until 2am updating the numbers.

None of these causes gets fixed by hiding the tool better. They get fixed by making the part that was always human, visible on purpose. Fourteen years inside Danaher’s commercial organization taught me that people don’t respect how you got to an answer. They respect whether you can defend it, and whether it drives the result. That’s the whole skill. The shame was never about the tool. It was about whether you knew how to use it well.