Imagine your team is asked to start using AI to draft customer responses.
The demonstration looks promising. Replies take seconds to produce, and the writing sounds professional. Then the questions begin. Who checks the information? What happens when a customer’s circumstances need more care? How much time will reviewing each draft take?
For the manager, introducing the tool means making decisions about how the work will change and helping colleagues understand what is expected of them.
CMI’s Leadership for AI programmes address this management responsibility, including leading change, applying governance and evaluating business value. CMI’s approach to leadership for AI
So where should a manager begin?
Choose a problem you can describe clearly
Before choosing a tool, write down what you want to improve.
In our customer-service example, the problem might be that staff spend too much time drafting routine replies. That gives you something specific to investigate.
You could trial AI-assisted drafting on a limited category of enquiries, using an approved tool and an agreed review process. Compare the results with your existing approach, including the time spent checking and correcting responses.
A useful trial should give you enough evidence to decide whether to continue, adjust the approach or stop.
Make room for questions from your team
Ask colleagues what they need to understand before the trial begins.
Someone may be worried about their role. Another person might already have spotted a weakness in the proposed process. Give both concerns a proper hearing.
Be clear about what has been decided and what is still open for discussion. If you do not yet know how responsibilities will change, say so and explain when you expect to revisit the question.
You can also involve staff in designing the trial. The people handling customer enquiries are well placed to identify awkward cases and suggest what a useful response should contain.
Keep judgement visible
An AI-generated recommendation needs the same scrutiny you would apply to other information used in a decision.
Suppose a tool suggests changing staffing levels because it predicts lower demand. Before acting, ask what information supports that forecast. Does it account for a forthcoming service change? Is there context your team knows that the system does not?
Agree who reviews recommendations and who has authority to act on them. For decisions with significant consequences, involve the relevant specialists and follow your organisation’s approval processes.
Managers should be able to explain the reasoning behind a decision, including how they used AI’s contribution.
Measure the work after the change
Decide what success would look like before the trial starts.
For AI-assisted customer replies, you might compare total handling time and the number of corrections needed. Ask staff whether the process helps them deal with enquiries, and check whether customers are receiving accurate, useful answers.
Include the effort required to maintain the process. A faster first draft may still need substantial checking.
Set a review date and record what you learn. That gives the next decision a firmer basis than enthusiasm about the demonstration.
Develop the skills your role now requires
If you are expected to lead AI adoption, consider where you need more support. It might be planning a project or helping your team work through an unfamiliar change.
Acudemy’s CMI Level 5 Diploma in Leading AI Transformation is designed for practising and aspiring middle managers with responsibilities in this area.
Its six selected units cover change leadership, ethical adoption, team confidence, AI-informed decisions, project management and leadership practice. Study is online, with a UK-based tutor supporting learners throughout and assessment through written assignments.
If these responsibilities are becoming part of your work, explore the Diploma and enquire with Acudemy. Bring a current workplace challenge to the conversation so the team can help you assess whether the course fits your needs.