ChatGPT (or AI) Hushing: The Last Resort to Justify our Employment?

“Can it be a crime if everyone does it?” That’s the question that popped into my head when I first started noticing how people were quietly using ChatGPT and other AI tools while pretending they weren’t. It’s like we’re all in on this secret, but nobody wants to admit it out loud.

The AI Washing to Hushing Shift

Remember the early days of AI hype? Companies were slapping “AI-powered” on everything like it was a magic sticker that made investors drool. We’ve all seen it: “AI washing,” as the folks at Corrs call it, where businesses inflate their AI capabilities to snag attention or cash (Corrs: AI washing and cyber washing). But here’s the twist, those overblown claims are starting to fizzle out. People aren’t biting as hard anymore.

What really counts now is actual value and user experience. Does the AI tool actually solve problems without making you want to pull your hair out? For the average person, the tech under the hood is about as interesting as watching paint dry. If a tool gets the job done efficiently, who cares if it’s AI or a team of elves?

But this shift from AI washing to AI hushing creates an interesting paradox. While companies were once exaggerating their AI use to impress, now we’re seeing the opposite: people quietly using AI tools while pretending they aren’t. It’s like we’ve collectively decided that admitting AI use somehow diminishes our human value.

With ChatGPT and AI tools, it feels like we’re jumping straight into the hushing phase. We’re not pretending AI doesn’t exist or that we’re using it when we’re not. Instead, we’re often just… not talking about it. At all.

The Social Desirability Bias Factor

This phenomenon isn’t new in the business world. In management accounting literature, there’s plenty of research about “performance overreporting” - where managers inflate their achievements to look better. But with AI, we’re seeing the opposite: underreporting of AI use. A recent SSRN paper by Yier Ling, Alex Kale, and Alex Imas explores this exact issue, calling it the “social desirability bias” - our tendency to present ourselves in ways that make us look good to others.

The paper, titled “Underreporting of AI Use: The Role of Social Desirability Bias” (September 2025), argues that we’re actively hiding our AI usage because it threatens our sense of human value. If AI can do the work, what does that say about us? Are we trying to justify our existence at work by pretending we’re still doing everything the “human” way?

I think that’s exactly what’s happening. There’s more at stake for employees than in previous workplace transformations. We need to feel like our contributions matter, like our expertise and creativity still count for something. Admitting we use AI feels like admitting defeat - like we’re not as smart or capable as we want others to believe.

Spotting and Hiding AI Use

Before we dive into the examples, let me be clear: I used AI to write this post and spend two hours customizing it. AI is such a powerful tool that can produce incredible outputs. But creating meaningful outcomes? That’s still very much in our hands. For beginners looking to navigate this space, here are a few things to watch out for. AI isn’t too complicated to spot if you know what to look for, but it’s also not something we should fear.

Examples of AI Hushing in Action

1. The Text Game

Let’s start with the most obvious: ChatGPT’s infamous em-dash usage. You know, those awkward dashes that appear in generated text —— it’s like the AI’s signature move. A Medium post by Brent Csutoras explores this signature quirk in detail.

How to spot it: Look for unnatural phrasing and over-polished language that feels too perfect. Tools like the Fluff-o-Meter can help detect AI-generated text quirks.

How to hide it: Manually edit out hyphens and add personal anecdotes or imperfections to make it feel more human.

But here’s the thing - is this really a problem? Most of us consume content on platforms like LinkedIn for entertainment and networking, not for deeply researched analysis. For that, we turn to other sources. A study published in The International Journal of Press/Politics about AI in news reporting found that while AI can generate content quickly, human oversight is crucial for accuracy and context.

2. Emoji Overload

Emojis have made a comeback thanks to ChatGPT. They’re everywhere in generated text, making everything feel a bit more… visual.

How to spot it: Excessive or awkwardly placed emojis that don’t quite fit the context. AI tends to sprinkle them liberally to add “personality.”

How to hide it: Use emojis sparingly and purposefully, or remove them entirely if they feel forced. However, they can actually enhance readability when used well.

Maybe emojis aren’t the enemy. They help break up text and add personality. In a world of endless scrolling, that visual element might be exactly what keeps people reading.

3. The Vibecoding Revolution

Then there’s the coding world. Tools like Lovable, Bolt, and V0 are revolutionizing web development. Using React and Next.js frameworks, they let anyone create professional-looking websites.

How to spot it: Websites that look remarkably similar in structure and styling, often with generic layouts and components.

How to hide it: Customize the generated code extensively, add unique features, and integrate your project with external services to differentiate your project. I am a vibe coder myself - check out some my projects like reference-analyzer.com that started with AI support but were heavily customized.

But here’s the catch: everything starts looking the same. For venture capitalists, this raises questions. Will startups built with “vibecoding” get discounted compared to those with senior developers? Probably, unless they show real traction. And let’s not forget security and data compliance - areas where many vibe coders have zero expertise.

4. Imagery and Authenticity

Finally, there’s the imagery debate. At what point can a company claim in their CSR report that they used real human photographers instead of AI-generated images?

How to spot it: Images that look too perfect or generic, lacking the authentic imperfections of real photography.

How to hide it: Clearly disclose AI use in reports, or combine AI-generated images with human-created content. A NZZ article explores how AI is changing creative professions in advertising.

For individuals, AI-generated content (often called “AI slop”) serves a purpose too. Its absurdity helps us slow down and appreciate real creativity. This Guardian article and another Guardian article have discussed this in more detail.

Creating a Culture of Clarity

So where does this leave us? AI hushing isn’t just about hiding our tools - it’s about preserving our sense of purpose. We’re in a transitional period where technology is reshaping how we work, and it’s scary. By not talking about AI, we’re trying to maintain the illusion that human creativity and expertise still reign supreme.

But maybe that’s the wrong approach. Instead of hushing, what if we embraced transparency? What if we celebrated how AI amplifies our capabilities rather than replaces them? The real value we bring isn’t in generating outputs - it’s in curating them, contextualizing them, and turning them into meaningful outcomes.

Most importantly, organizations need to create clear cultures around AI use. If AI is prohibited due to confidentiality or data sensitivity concerns, usage should be monitored and sanctioned when rules are broken. However, if your environment has no downside to AI use, it’s crucial to communicate an open AI culture so people can speak freely about their AI usage instead of having to hide it. This transparency reduces the social desirability bias and allows everyone to focus on what really matters: delivering value through human-AI collaboration.

As we navigate this AI-driven world, let’s remember: the tools don’t define us. Our choices do. And choosing to be honest about how we work might just be the most human thing we can do.

What do you think? Are you hushing your AI usage at work, or are you being open about it? I’d love to hear your perspective…