Making AI Adoption Work for the People Doing the Work

AI adoption is often discussed in terms of productivity, efficiency and competitive advantage. Organisations invest in tools, redesign processes and encourage employees to become more productive.

But are we giving enough consideration to the people who are expected to make this transformation happen?

We ask employees to share their knowledge, identify automation opportunities, and learn new ways of working. At the same time, many are uncertain about what these improvements mean for their jobs, responsibilities and future.

If employees believe that making AI more effective could put their own jobs at risk, why would they actively contribute to that change?

Before discussing how to increase AI adoption, we need to understand what concerns employees.

Understanding employee resistance

Resistance to AI is not always about an unwillingness to change or learn. Employees may recognise its benefits and still hesitate to use it.

There are several reasons for this.

1. Fear of losing the job

An employee who has spent years developing expertise in a business process is now asked to share that knowledge to help automate it.

From the organisation's perspective, this improves efficiency. From the employee's perspective, it raises a simple question: What happens to me once the process no longer needs my involvement?

This concern is not limited to repetitive jobs. Experienced professionals may also question how AI will affect the value of their expertise.

Employees need to understand how their roles and opportunities could evolve, rather than being expected to contribute to automation without knowing what comes next.

2. Faster execution, more work

AI reduces the time required to complete many tasks. But what happens when organisations adjust their expectations to match that speed?

A task that previously took two days may now take a few hours. Soon, that improvement becomes the new baseline, and employees are expected to complete more work within the same time.

AI may reduce the effort required to create something, but reviewing, validating, and making decisions still take the same amount of time.

Instead of reducing the workload, AI can shift the pressure from creation to evaluation.

If every productivity improvement adds work and compresses timelines, employees may see AI as another burden rather than something that makes their working lives easier.

3. Pressure to maintain quality and speed

AI can generate results quickly, but faster generation does not guarantee better outcomes.

Employees are expected to deliver at AI-enabled speed while maintaining the same standards of quality and accountability.

A product manager may generate several product concepts in a day, but evaluating customer needs and commercial viability still takes time. Similarly, an HR professional may prepare communications or policies faster, but the implications of those decisions still require human judgement.

When timelines are based on how quickly AI generates output rather than the time required to evaluate it, employees face unnecessary pressure.

The organisation benefits from faster execution, while the employee remains responsible for the outcome.

4. Resistance to learning

Not every employee is equally comfortable with technology.

An experienced employee may have deep business knowledge and strong customer relationships but feel uncomfortable when asked to adopt unfamiliar AI tools.

Some may worry about looking inexperienced in front of colleagues. Others may struggle to find time to learn while continuing to meet their existing responsibilities.

We should distinguish between resistance to learning and resistance to how we introduce learning.

Employees do not necessarily need to become technology experts to use AI effectively. They need to understand where it can help in their work, how to evaluate its outputs and when their own judgement is essential.

Their existing experience should remain valuable. AI should help them apply that knowledge in new ways rather than make them feel that their experience is becoming irrelevant.

5. Fear of making mistakes

Employees may respond to AI in two opposite ways.

Some may rely on AI too heavily, accepting its outputs without sufficient verification, particularly when they are under pressure to deliver faster.

Others may avoid AI altogether because they are concerned about accuracy, accountability or the consequences of making a mistake.

Both reactions can come from the same uncertainty.

Employees may be encouraged to use AI but criticised when an AI-assisted output contains an error. Equally, they may be criticised for working too slowly because they spent additional time validating the results.

How do employees know when to trust AI, when to challenge it, and what happens if they make the wrong decision?

Without clear expectations, employees may feel that they are taking a personal risk regardless of how they use the technology.

Confidence cannot be created through communication alone

Organisations often reassure employees that AI is intended to augment people rather than replace them.

But employees will judge that message against what actually happens.

If AI adoption is followed by reduced staffing, increasing workloads or unrealistic delivery expectations, reassurance alone will not build confidence.

At the same time, leaders cannot promise that every job will remain unchanged. AI will change responsibilities, eliminate certain tasks and potentially reduce the need for some roles.

Pretending otherwise does not create lasting trust.

Organisations need to communicate what has been decided, what remains uncertain and how employees will be supported through the transition.

Employees do not necessarily need a promise that nothing will change. They need confidence that the organisation will treat them fairly and consider their future as part of the transformation.

Make the benefits visible to employees

We often hear that AI will free employees from repetitive work so they can focus on more valuable activities.

But what exactly are those activities?

If AI reduces the time an employee spends on routine tasks, can they now focus on improving customer service, solving more complex problems or developing new capabilities?

Can the additional capacity help them manage their workload better rather than simply increasing the number of tasks they are expected to complete?

Organisations need to make these opportunities visible.

When an employee helps improve productivity, we should not measure the benefits only through cost savings or increased output. We should also consider what improves for that employee.

Not every hour saved by AI needs to be filled with additional work. Sometimes the benefit is simply having more time to do existing work properly.

Involve employees in shaping the change

Employees understand the practical realities of their work. They know where processes fail, which exceptions require attention and what customers actually expect.

Their knowledge is essential for identifying where AI can genuinely improve operations.

However, participation should not mean simply asking employees to document their knowledge so that existing activities can be automated.

Employees should have a voice in deciding what to automate, how AI-supported processes are designed and how their responsibilities will evolve.

Their feedback should also help evaluate whether AI is delivering the expected improvements.

If an employee reports that AI has reduced preparation time but increased validation effort, that feedback should be treated as valuable operational evidence rather than resistance to adoption.

Their contribution should also be recognised through development opportunities, performance discussions and appropriate rewards.

Employees are more likely to participate when they can see that their knowledge and contribution remain valuable.

Give employees room to learn safely

In our earlier article, AI Adoption Is Not Learning, we discussed why introducing AI tools is not the same as building the capabilities needed to use them effectively.

But learning also requires confidence.

Employees need time to experiment, ask questions and make mistakes without feeling that their professional capabilities are being judged.

Training should be relevant to their work and should not assume that everyone has the same technical background.

Managers should encourage employees to question AI outputs and report errors rather than treating such behaviour as resistance.

An employee who identifies an AI-generated mistake may be preventing a much larger problem.

The objective should not be to make everyone use AI as much as possible. It should be to help employees understand when and how to use it responsibly.

Leadership behaviour matters more than reassurance

An organisation cannot promise better working conditions while turning every productivity improvement into a more demanding delivery target.

It cannot encourage responsible AI use while rewarding speed alone and penalising employees who take the time to validate results.

Similarly, it cannot promise career development without giving employees the time and opportunities to develop new capabilities.

Leadership decisions need to support the message employees hear.

When setting targets, we should consider the full effort required to deliver an outcome, not just how quickly AI can generate an output.

When redesigning roles, we should consider how employees can apply their existing knowledge in new responsibilities.

When employees raise concerns about workload, quality, or the impact of AI on their roles, we should treat those concerns as input to improving the transformation.

AI adoption should not become a target that must be achieved regardless of its impact on the people doing the work.

A more productive future should work for employees too

AI can help organisations serve more customers, improve quality, develop new products and create new business opportunities.

It can also help employees reduce repetitive work, develop new capabilities and focus on activities where their experience and judgement make a difference.

But these benefits will not happen automatically.

If organisations measure AI adoption only through cost savings, reduced staffing requirements or increased output, they may overlook the people responsible for achieving those improvements.

We need to consider employee confidence, workload, learning and opportunities alongside organisational productivity.

Employees will need to adapt as AI changes how work is performed. But organisations also have a responsibility to create an environment in which employees can participate confidently in that change.

AI adoption should not leave employees feeling that the better they become at using the technology, the less secure their future becomes.

When we ask employees to help build a more productive organisation, we should be able to answer one simple question:

What does a more productive future mean for the people doing the work?