Should You Tell Your Boss You're Using AI at Work?

Ellie at Willa, Editorial lead, Willa · Published September 29, 2026 · Updated September 29, 2026

Somewhere around 80 percent of workers now use AI on the job, according to Cornerstone OnDemand's October 2025 survey of workers in the US and UK. Here is the part that should get more attention: 57 percent of the US workers surveyed said they were reluctant to tell their manager or colleagues about it.

That is a lot of people doing good work quietly, wondering if the thing making them faster is also the thing that could get them flagged. If you have ever finished a task in an hour that used to take an afternoon and then sat there deciding whether to mention how, this article is for you. The short version: telling your boss is almost always the right move, but there is a right way to do it, and a few things you should sort out first.

Why do so many people hide their AI use at work?

The fear is not irrational. In a Cox Business survey of more than 1,000 Gen Z and millennial workers, 47 percent said they feared AI could replace their jobs, and that fear was a top reason they kept their AI use to themselves. The logic goes: if I show them the machine can do half my job, why would they keep paying me for all of it?

There is also plain confusion. In that same survey, 30 percent of workers said they were either unfamiliar with their company's AI policy or their company did not have one at all. When the rules are unclear, silence feels safer than asking a question that might draw attention.

And the quiet is contagious. ResumeBuilder's April 2026 survey of 1,000 full-time US employees found that 62.8 percent of people who used AI to handle tasks a coworker used to do never told their manager. At companies that had already been through layoffs, that number climbed to 74.3 percent. Fear of layoffs makes people hide their AI use, and hidden AI use makes it easier for leadership to misjudge what their team actually does all day. Nobody wins that cycle.

Can you get in trouble for using ChatGPT at work?

You can, but almost never for the reason people worry about. The realistic risks are specific, and they are worth naming plainly.

The first is data exposure. Pasting confidential material into a consumer AI tool can violate your employment agreement, client contracts, or privacy law, depending on what you paste. This is the risk companies care most about, and it is the one that turns "she used ChatGPT" into an actual incident.

The second is policy violation. If your company has a written AI policy and you work around it, the tool is not the problem; the workaround is. Even a policy you think is too strict is still the policy until you get it changed.

The third is unreviewed output. If AI writes something wrong, biased, or invented, and you ship it with your name on it, the mistake is yours. "The AI did it" has never once worked as a defense, and it never will.

Notice what is not on this list: being more productive. No one gets fired for finishing the report early. People get in trouble for what they fed the tool and for what they failed to check on the way out.

What should you never put into an AI tool at work?

Before any conversation with your boss, get your own habits clean. Unless your company provides an enterprise AI tool that is explicitly approved for sensitive data, keep these out of your prompts:

1. Customer and client information. Names, contact details, account data, anything covered by a privacy commitment your company has made.
2. Confidential business material. Financials, unreleased plans, contracts, legal matters, anything marked internal.
3. Credentials and security details. Passwords, API keys, system configurations, and anything about how your company's infrastructure works.

A useful gut check: if you would not email it to someone outside the company, do not paste it into a personal AI account. Most of the value of AI at work comes from formatting, drafting, summarizing, and thinking through problems, and you can do all of that with the sensitive details stripped out or swapped for placeholders.

How do you tell your boss you use AI without making it weird?

Frame it around outcomes, not tools. Your boss cares about the quality and speed of your work; the software behind it is a supporting detail. A version of this works in most jobs:

"I've been using AI to handle the first draft of [specific task], which has been saving me around [amount of time] a week. I review everything before it goes out. I wanted to flag it so I'm following whatever guidelines we have, and because I think a couple of these workflows could help the rest of the team."

That short script does three things. It anchors on a result rather than a confession. It shows you have a verification habit, which is the thing managers actually worry about. And it positions you as someone thinking about the team, which is how you go from "person who uses AI" to "person we ask about AI." That last identity is worth real career capital: the Cornerstone survey found only 44 percent of US employees have received any AI training, and 65 percent want it. The gap between those numbers is an opening for whoever is willing to be visibly good at this.

If your work has an obvious fingerprint problem, fix that before you disclose anything. AI drafts that sound like a press release will undermine your "I review everything" claim. We wrote a whole guide on making AI write like you, not a robot, and it pairs well with this conversation.

What if your company has no AI policy?

Then you have an opportunity dressed up as a gap. With roughly a third of workers unsure whether a policy even exists, being the person who asks is a leadership move, not an admission.

Keep the proposal small. Ask your manager whether there is guidance on AI tools, and if there is not, offer to draft a one-page starting point: which tools are fine, what data stays out of them, and who reviews AI-assisted work before it ships. You do not need to be technical to write that page. You need to be the person who noticed it was missing.

This also changes your posture from asking permission to setting terms. A policy written with input from people who actually use the tools daily will be more realistic than one written in a conference room after an incident. Volunteering early is how you make sure the eventual rules are ones you can live with.

One more place this pays off: interviews. Employers now routinely ask candidates how they work with AI, and "I used it quietly and hoped nobody asked" is a weak answer. If you can say you disclosed it, verified outputs, and helped shape your team's guidelines, you have a story. Our guide to answering "How do you use AI?" in a job interview shows how to turn that story into an offer.

Key takeaways

• Disclose in most cases. Hidden AI use protects you less than it feels like it does, and the data shows most of your coworkers are hiding it too.
• Read the policy before the conversation. If your company has AI rules, follow them first and push to change them second.
• Keep sensitive data out of personal AI accounts. Data exposure is the risk employers actually punish, not productivity.
• Frame disclosure around results and review. "It saves me three hours a week and I check everything" answers the question before it is asked.
• If there is no policy, volunteer to draft one. The person who writes the first page usually shapes everything that follows.

Most of us figured out AI at work alone, in a browser tab we were not sure we should have open. It goes faster together. Willa hosts expert-led sessions almost every week for women building careers with AI, covering exactly these kinds of real-workplace questions. See what's coming up on our events page, or become a member and start comparing notes with women who have already had this conversation with their boss.

• ai at work
• workplace ai policy
• ai transparency
• career development
• ai skills

Join Willa, the community where women learn AI together.

All posts