Creating Psychological Safety Around AI Use · AI for Managers

Fear of looking foolish, using the tool wrong, or producing bad work is the primary barrier, often below the surface. Managers control whether failure in front of the team is safe.

Fear of looking foolish is rational.

People also experience gen AI as a threat to competence, autonomy, and relatedness. "Just use Copilot more" lands as "your judgment is optional." Safety is built by what you do: share your own bad outputs, invite "I pushed back on the model," respond to failure with curiosity ("what did the prompt need?"), name learning in public misses.

Core principles

Job concerns: "less valuable" vs "redundant" need different responses. AI handles volume; judgment, relationships, and accountability stay human. Say that out loud. Do not joke that Copilot will replace the team.
Invite pushback as a professional move: "What did you refuse to send?" is a better standup question than "How many prompts this week?"
Be direct about job replacement fear; connect AI to removing friction so more judgment work is possible.
Do not perform vulnerability, actually share a weak first draft and what you fixed.

Check yourself

What does this lesson identify as the primary barrier to AI adoption on most teams?

Do this in Copilot

Share one honest read on a recent team AI attempt, what you expected, what happened, what you would do differently.

Paste this into Copilot Chat and work through it before moving on.

Team conversation starter

Draft three open questions I can ask my team in our next meeting to surface honest AI experiences, including times it did not help and times someone pushed back on the model, without making people feel evaluated.

Techniques: Psychological safety

Open Copilot

Did you run this in Copilot? Mark complete when you have tried it.