Risk review and bringing worried staff along

When you finish, you can review AI risks before launch and respond to staff concerns respectfully.

A simple risk review

Before using AI on a task, ask: What could go wrong? Who could be hurt? What data is involved? Is there a rule or law about this? Who checks the output? What is the plan if it is wrong? Write the answers down.

If the task involves private data, safety, health, legal duties or decisions about people, ask your privacy, legal or compliance contacts first. In regulated work, the licensed professional decides and the AI only helps.

Common risks and fixes

Wrong facts: check against sources. Private data leaking: use approved tools and remove identifiers. Bias: review wording and outcomes for fairness. Over-reliance: keep people skilled by doing some tasks without AI. Unclear ownership: name a person responsible for each use.

Add a way to report problems without blame. Mistakes found early are cheap. Mistakes hidden are expensive.

Listen to worried staff

Some people fear losing their job. Others worry about quality, fairness or being watched. Their worries are reasonable. Begin by listening. Say what you know and what you do not know. Do not promise what you cannot control, and do not wave concerns away.

Offer real things: training time, a chance to shape the pilot, and a clear statement of how you will and will not use AI on your team, such as removing dull tasks rather than judging people. Be honest that some tasks will change.

A worked example

At a school office in Nairobi, a clerk named Grace says, "I'm afraid this will replace me." Her manager, Daniel, listens, then answers: "I can't promise how work will change, but this pilot is for drafting letters, and you know our families best. I'd like you to lead the checking."

Grace joins the pilot, finds the tool helps with routine letters, and becomes the team's go-to person for what to check. Respect turned fear into leadership.

Practice now

Run the risk review questions on one planned use of AI and write short answers.
Hold a 15-minute listening session where each team member can name one hope and one worry about AI.

Use any AI assistant you have. Do not paste private, client or patient information into it.

Check yourself

1. What is the first step with a worried team member?

2. Which task needs extra caution and expert review?

3. Why allow blame-free reporting of problems?

Ask the AI tutor

Ask anything about this lesson, in any language. The tutor is an AI: it can be wrong, and it never needs your personal details.

Take the final exam

AI-drafted. Checked by an automated validator and a separate AI reviewer; a person has not read every lesson. Translations and tutor answers are AI-made. Found a mistake? Tell us. AI can be wrong. General information only, not professional advice. Not affiliated with any AI company named. Current University is not accredited.