How to control AI use: Your team is using AI. The only question is: Are they hiding it from you?
Microsoft and LinkedIn’s 2024 Work Trend Index found that 75% of knowledge workers use generative AI at work, and 78% of AI users are bringing their own tools (BYOAI), often because official guidance is missing.
In many organizations, AI use is already happening unevenly (some teams are using it openly, some secretly, some not at all). Meanwhile, leaders feel more and more pressure to do something about it.
Before you can set good rules, you need to see what’s actually happening. That means changing how you talk about AI use so people feel safe being honest with you.
Visibility is the foundation of any effective AI strategy for leaders.
This article includes an AI Visibility Statement you can say to your team to make transparency the norm and reduce hidden risk.
Why people hide AI use:
The real risk isn’t that people are using AI. It’s that they’re using it without clear expectations and without feeling safe to tell you about it.
When people worry they’ll be judged, they hide what they’re doing. And when you can’t see what’s happening, you can’t guide it.
Most “shadow AI” is not malicious
In nearly every case I’ve seen, shadow AI doesn’t come from recklessness. It comes from capable people trying to save time and get work done faster.
People mean well when they use AI, but unseen use can put your organization at risk. As a leader, you need to know what’s happening so you can spot risk and create clear guardrails and policy.
If your leadership team hasn’t defined what’s acceptable, what’s off-limits, and what good use looks like, your team will make their own decisions. When that happens, people start using AI in inconsistent ways. Not because they’re careless, but because leaders haven’t made their expectations clear.
What visibility is, and what it is not
Leaders often worry that visibility means surveillance. It doesn’t.
Visibility is not surveillance.
You are not asking people to log every prompt. You are not monitoring individuals.
Visibility is shared awareness of systems and risks.
You need to know how AI use affects:
- Client deliverables or client communications
- Sensitive data or confidential information
- Decisions with legal, financial, safety, or reputational impact
- Work that sets direction, commitments, pricing, policy, or hiring outcomes
Here’s how to explain it to your team:
We’re not asking for perfect AI use. We’re asking for visible AI use.
The visibility paradox
Many leaders ask how to control AI use in their organization. The instinct is to tighten rules. But when employees want to be efficient, they experiment with AI anyway. This tension can create risk when it goes unaddressed.
- If someone isn’t sure whether AI use will be seen as “smart” or “lazy,” they won’t mention it.
- If someone isn’t sure whether they’ll get in trouble for using a public tool, they won’t ask first.
- If someone worries their competence will be questioned, they’ll keep it private.
That means the organization loses the one ingredient readiness depends on: visibility.
The more leaders clamp down without offering a safe alternative, the more usage goes underground.
What to say to your team (the AI Visibility Statement)
To get people to tell you about AI use, you need to make it clear they won’t get in trouble for being honest.
Use this AI Visibility Statement with your team to get them talking about AI use:
“We’re open to AI, and we’re serious about using it responsibly. If you use AI on work, please disclose it, you won’t get in trouble for being transparent. That visibility helps us protect sensitive information and agree on standards we can trust.”
It does three things immediately.
1) It lowers fear: People stop assuming disclosure will be punished. The intent is explicit. No gotcha.
2) It creates a learning culture: The goal shifts from catching mistakes to building shared judgment, practical habits, and clear standards.
3) It forces leadership to lead: You cannot govern what you refuse to see. Visibility becomes the foundation for responsible adoption.
What happened when one leader asked for visibility
A leader I worked with assumed people weren’t using AI much because no one talked about it.
After sharing the AI Visibility Statement with this leader, they added one question to weekly team meetings: “Where did AI support your work this week, and what did you verify?”
Within two weeks, three things came up. People were using AI to draft client emails and summarize research. One person also shared an uncomfortable moment where they pasted information into AI that shouldn’t have been shared.
Nothing exploded. No one was shamed. The leader thanked the person for being honest, explained the boundary clearly, and created a simple rule: all client-facing work needs a second review.
The result wasn’t perfect, but people stopped guessing what was okay and started making better decisions.
This is what AI readiness looks like in action.
How to create visibility without creating fear
The AI Visibility Statement is just the first step. You also need a simple system that makes it easy for people to talk about AI use regularly.
Step 1: Share the statement
Use the AI Visibility Statement from earlier to tell your team that being honest is safe and that you’re creating clear rules.
Step 2: Create regular check-ins
Choose one regular time to ask about AI use.
For example:
- A monthly leadership check-in: “Where did AI show up in work this month?”
- A team retro prompt: “What worked well with AI? What felt risky? What should become a standard?”
- A manager 1:1 question: “Did AI support any of your work this week? What did you verify?”
The goal is to make these conversations normal, not to judge performance.
Step 3: Give people an easy way to share
People need a simple place to share what they’re doing and ask questions.
For example:
- A shared Slack or Teams channel
- A simple email thread
- A recurring agenda item in team meetings
Keep it simple. You’re building visibility, not adding paperwork.
Step 4: Turn what you learn into action
Once a month, write down what you learned. Keep it to one page.
The Readiness One-Pager
- Known good uses: what you want repeated
- Red lines: what never goes into AI
- Review required: work that needs a second set of eyes
- Open questions: what you are still clarifying
This creates a record that prevents the same questions from coming up every month.
The manager playbook. Five scripts that make visibility safe
Managers are the key to making visibility work. Give them these five questions that help people think through their AI use without making them feel policed.
Use these five questions:
- “Show me how you used AI. What did you keep, and what did you change?”
- “What did you verify before sharing this externally?”
- “What part of this task should never be automated, even if AI can?”
- “Is this a ‘review required’ situation? If yes, who should review it?”
- “Do we need a team standard here so people do not have to guess?”
These questions help people build better judgment about when and how to use AI.
What if you discover risky AI use
If someone tells you they did something risky with AI, here’s how to respond in a way that keeps trust and reduces harm:
- Thank them for transparency. “I appreciate you telling me.”
- Separate learning from discipline. “Our goal is to learn and set clear boundaries.”
- Address the specific risk quickly. Remove or correct what needs correcting.
- Clarify the boundary in one sentence. Make it memorable.
- Share the learning without naming the person. Turn it into a standard, not a story.
When leaders respond this way, people learn that being honest leads to clarity, not punishment.
A note for regulated and high-confidentiality environments
If you work in health, finance, legal, government, or other regulated industries, visibility is even more important and boundaries need to be stricter.
In these industries:
- Use approved tools only
- Remove or hide sensitive information by default
- Create “review required” rules early
- Be clear about what information can never go into AI
You can still create a culture where people learn and improve. You just need clearer rules about what’s allowed.
What to do this week
If you do nothing else, do this:
- Share the AI Visibility Statement
- Add one regular time to ask about AI use (like a weekly check-in or team meeting question).
- Ask leaders and managers to share their own AI use first.
When leaders are honest about their own AI use, teams feel safer doing the same.
Ready to build AI readiness across your organization?
Sharing the AI Visibility Statement is a strong first step. But leaders need more support to move from scattered AI use to clear standards and confident decision-making.
The AI Readiness Accelerator for Leaders is designed for founders and leadership teams who know AI matters, feel the pressure to act, and want clarity before making big decisions.
In this program, you’ll:
- Build visibility into AI use without creating fear
- Set clear boundaries that protect sensitive information and reduce risk
- Create standards that guide AI use instead of letting people guess
- Build confidence in your leadership team to make decisions about AI that align with your values
This isn’t about picking tools or automating processes. It’s about building the readiness your organization needs before scaling AI use.
