Using ChatGPT at Work Could Hurt Trust, Promotions, and Pay

What the Research Actually Found

The team ran six studies examining how LLM use shapes workplace evaluations. One experiment asked participants to act as supervisors and decide whether they’d trust an employee’s reported work hours — or verify them independently.

When AI use was involved, supervisors were more likely to verify. That’s not a small signal. Verification behavior is a direct proxy for distrust.

Managers also rated AI-assisted work as lower quality — even when the output itself wasn’t objectively worse. The stigma appears to attach to the process, not just the product. Related concerns about how AI tools shape professional judgment appear elsewhere too.

The Hidden Career Tax

This creates a real and underappreciated tradeoff for workers trying to be “AI-first”:

• Performance reviews may be quietly downgraded when AI use is known or suspected.
• Promotions could stall if managers associate AI reliance with lower capability or effort.
• Pay increases tied to perceived contribution may shrink when that contribution is attributed to a tool rather than the person.

None of this shows up in a job description or an AI policy memo. It’s the informal layer of evaluation that shapes careers.

Why Managers React This Way

The reaction isn’t entirely irrational. Managers are trying to assess who is capable, not just what got done. When a tool does the heavy lifting, the signal about the employee’s skill becomes noisier.

There’s also a trust-effort link at play. If a manager believes you offloaded cognitive work to a chatbot, they may question how much you actually invested — regardless of the output quality.

This is a perception problem as much as a performance problem, especially in broader discussions about AI in the workplace.

The Visibility Trap

Part of what makes this tricky is that AI use is increasingly visible. Paste something into ChatGPT, get a polished paragraph back, submit it — and a pattern emerges over time that colleagues and managers notice.

Workers who use AI quietly and integrate it into a clearly skilled workflow may avoid the penalty. Workers who lean on it visibly, or whose output suddenly shifts in tone and quality, may not.

The tool isn’t the problem. The signal it sends is.

What This Means If You’re Using AI at Work

You don’t have to stop using AI tools. But you do need to think about how your use is perceived — and whether you’re building or eroding the trust that actually drives career outcomes.

A few practical framings worth considering:

• Show your thinking, not just your output. If you used AI to draft something, add your own visible layer of judgment, editing, or strategic framing.
• Don’t let AI flatten your voice. Generic, polished-but-bland output is a tell. Managers recognize it.
• Understand your manager’s mental model. Some leaders see AI use as resourcefulness. Others see it as a shortcut. Knowing which you’re dealing with matters.

The smartest use of AI tools at work right now isn’t just about efficiency. It’s about using it in ways that preserve — or strengthen — the trust that promotions and pay actually run on.