Skip to content

Address team questions and concerns

Concerns are a natural response

Resistance to AI is not necessarily irrational. Change can affect employees’ identity, competencies and sense of security. Understanding the source of a concern is the first step towards addressing it constructively.

Common sources of concern

1. Fear of losing a job

What you may hear: “AI will replace us.”

What to establish: The effect depends on the task, role and organisation’s decisions. AI can change job content, required skills and the number of positions. Without a specific plan, you cannot promise that no one will be replaced.

How to respond:

  • Explain how particular roles may change, rather than relying on general reassurance.
  • Share verified examples of AI supporting work and openly discuss possible employment effects.
  • Invest in retraining and skills development.

2. Distrust of AI outputs

What you may hear: “AI invents things; we cannot trust it.”

What to establish: AI can generate incorrect information and convincing but false explanations. Assess the risk on specific tasks and consider the consequences of an error.

How to respond:

  • Teach critical evaluation, the fourth competency of the AI-Native individual in L2 — Transform.
  • Establish review and validation processes.
  • Share cases where AI helped find mistakes, with evidence and limitations.

3. Loss of control or autonomy

What you may hear: “I do not want AI telling me what to do.”

What to establish: AI may affect work allocation and performance assessment. Explain the system’s actual authority, oversight and the process for challenging or correcting an output.

How to respond:

  • Define who decides, when human review takes place and who can stop the system’s use.
  • Clarify where people have a choice and which procedures are mandatory; do not promise autonomy the process does not allow.
  • Provide ways to give feedback and improve workflows.

4. Uncertainty about data privacy

What you may hear: “Where does our data go?”

What to establish: This is a legitimate concern. Without clear rules, employees may share sensitive data with unverified tools.

How to respond:

  • Establish an AI policy, supported by L1 — Align.
  • Use approved tools with transparent conditions.
  • Explain which data may be shared, for which purpose and under what conditions.

Working constructively with concerns

Involve people in designing the change

Instead of announcing “Everyone will now use AI,” invite participation around a concrete problem: “We want to test AI on problem X. Who would like to join the pilot?”

Involve sceptics in evaluation

People asking critical questions can help identify real risks. Invite them to review proposed uses and the pilot’s acceptance criteria.

Build trust through measured results

Share a measured result early in the pilot, including its limitations. A claim such as “AI saved two hours per week on reporting” needs evidence, a defined sample and time spent on review. An unsuccessful pilot is also useful evidence.

Keep communicating

  • What AI will change and what remains uncertain.
  • The results of pilot projects.
  • How emerging problems are being addressed.

Signals for discussion and improvement

Low adoption or training attendance does not by itself establish resistance. Causes may include an unsuitable tool, lack of time, access barriers or no relevant task. Interpret metrics in context rather than automatically rating individuals.

IndicatorHow to observe it
AdoptionProportion of employees actively using approved tools
SentimentSurveys about experiences and attitudes, for example quarterly
Shadow AIIdentified unauthorised AI tools in the organisation
Learning participationAttendance at voluntary AI learning activities

Next steps