How to Talk to Your Manager About Using AI

Start by asking what's approved

"Can I use AI?" puts your manager on the spot and invites a quick no. A better opener asks how to do it the right way. Any manager is glad to hear that question.

Your company's AI policy, sometimes called an acceptable use policy, should answer four things. They're the same four questions from Is It Safe to Use AI at Work?, and here they become your opener. Ask about any you can't answer from memory:

1. Which AI tools and connectors are approved?

2. Which account should I use for customer data?

   A company account, or is personal ever OK?

3. What's off limits?

   Some uses may be restricted, such as pricing, contracts or recording calls.

4. Who do I ask when I'm not sure?

If there's no policy yet, you've just helped your manager notice. Offer to stick to public information until there is one. Managers who need to write one can start with AI Policy for Sales Teams.

Propose a small, measured experiment

"AI could save us so much time" sounds like hype. One task with a number attached sounds like a plan. The method is simple: time a task by hand first, then time it again with AI.

1. Pick one task

   Choose something you do every week that's repetitive, mostly text and easy to check. Researching an account before a first call or cleaning up call notes are good candidates. Writing pricing proposals is not.

2. Time it by hand first

   Do the task the usual way three times and write down the minutes each time. That's your baseline.

3. Name the guardrails

   Which approved tool, which data (see the data traffic light) and how you'll check the output before anyone else sees it.

4. Run it for two weeks

   Use AI on that one task only. Time each one, the same way.

5. Report back

   Before, after, what went wrong and whether it's worth keeping.

Haven't tried AI on a real task yet? Do Your First Hour with AI at Work before you propose anything, so you know which task to pick.

Keep your time log this simple. Fill in your own numbers. Don't borrow anyone else's.

Share the win

Report back after two weeks, even if the result is mixed. Honest reports earn the next experiment.

The last line matters most. Saved minutes only count if they turn into selling.

Already past your first experiment? Your AI Plan: AI Power Seller shows how to turn this update into a monthly impact note.

When your manager is skeptical

A skeptical manager usually isn't against AI. They're protecting something: customer data, accuracy, buyer relationships or the team's time. Find the worry and answer that.

The worry about made up facts is real. AI can state a wrong fact with total confidence; that's called a hallucination. Agreeing with a fair concern makes you more persuasive. If your manager believes a myth, such as "AI just searches the internet," 10 Myths About AI in Sales has plain corrections you can share. What Is AI? is a good five minute read for a manager starting from zero. If the worry is buyers spotting AI, How to Tell If Something Was Written by AI shows the edits that prevent it. If the worry is replacement, What AI Can't Do in Sales (Yet) is an honest look at the parts of the job that stay with people.

Prepare the conversation with AI

You can use your company approved AI tool to get ready. Fill in the brackets with your own details, then edit what comes back so it sounds like you.

I'm a prospecting rep. I want a 10 minute conversation with my sales manager in our next one on one about using AI at work. I'd like to ask what's approved and propose one small experiment: using our company approved AI tool to [task], timing it before and after for two weeks.

My manager cares most about [pipeline / quality / not wasting time] and has said [anything they've said about AI].

Help me prepare:
1. A one sentence opener
2. My four questions about what's approved
3. A two sentence proposal for the experiment
4. The three objections my manager is most likely to raise, with a calm answer to each
5. One line to close with a clear next step

Plain language. No hype words.

Then practice out loud. This works the same way as practicing a cold call.

Act as my sales manager. You're fair but skeptical about AI: you worry about customer data, made up facts and reps wasting time on new tools. I'll propose a two week experiment using AI for [task]. Push back the way a busy manager would, one objection at a time, and wait for my answer each time. After four rounds, step out of the role and tell me which of my answers were strongest and which sounded vague.

After two weeks, let AI tidy up your report, but only from your own numbers.

Here's my time log for [task] before and after using AI for two weeks:

[paste your own time log]

Write a four line update for my manager: the task, the time before and after, what I checked to keep quality up, and what I did with the time I got back. Use only the numbers in my log. Don't round up or add anything.

Optional: suggest it for the team

If your experiment worked, your manager may ask, "How do we get the rest of the team doing this?" Start with the cheapest options:

• A 15 minute show and tell in a team meeting, where you walk through one task live
• A shared document of prompts that worked, so nobody starts from a blank page
• A standing question in pipeline reviews: "What did AI help with this week?"

Some teams want more structure, such as a shared seller context file for everyone, common rules and practice on real accounts. That's when training comes up. If you suggest it, keep it short and let your manager decide.

Quick check

FAQ