K3 Insights

"The fastest-growing source of workplace stress in 2026 isn't AI itself - it's the fear around it."

That is how a recent Human Resources Director article describes the view of Seemali Shukla, HP's People Leader for Australia and New Zealand. It is a useful distinction. AI may be the catalyst for workplace anxiety, but the questions employees are asking are familiar: What does this mean for my role? Will I be able to keep pace? Am I being supported - or monitored? Is today's productivity tool tomorrow's restructuring rationale?

AI may be new. The fundamentals of managing workplace change are not.

Fear, resistance, inconsistent uptake and loss of trust are familiar outcomes when change is introduced without a clear purpose, meaningful employee involvement, honest communication or appropriate support. Employers do not necessarily need an entirely new model for managing AI. They need to apply established change-management practices thoughtfully to the particular capabilities and risks of AI.

Treat AI adoption as a people change

It can be tempting to approach AI as a technology project: select a platform, establish access and train employees to use it. That may implement the tool, but it does not necessarily achieve useful or sustainable adoption.

Employees are unlikely to assess AI only by what the tool can do. They will also interpret what its introduction says about the organisation's intentions, the future value of their role, how performance may be measured and whether they will be supported through the transition.

Resistance should not automatically be characterised as an unwillingness to innovate. It may indicate that employees do not understand the purpose of the change, do not trust the stated rationale, cannot see how the tool fits their work, or have legitimate concerns that have not yet been addressed.

Start with the problem, not the product
Before introducing AI, employers should be able to explain the workplace problem they are trying to solve. "Because AI is the future" is not a meaningful change rationale.

A clearer purpose might be to reduce repetitive administration, improve access to information, shorten the time spent navigating multiple systems, or allow employees to focus more attention on customers, relationships and complex decisions.

The purpose should be specific enough to evaluate later. If a tool is intended to reduce cognitive load, employers should assess whether it has actually made work more manageable - not simply whether output has increased.

Be honest about what is - and is not - known

AI is developing quickly and employers may not be able to predict every effect. Employees do not necessarily require certainty; they do require candour.

Good communication distinguishes between:

• what has already been decided;
• what is being tested or piloted;
• what remains uncertain;
• what information will inform later decisions; and
• when employees can expect further updates.

This is particularly important if AI could affect role content, work allocation, performance expectations or future workforce requirements. Reassurance that cannot be substantiated may calm concern briefly, but it can damage trust later.

Involve the people who perform the work
Employees are often best placed to identify which parts of a process create unnecessary cognitive or administrative load. They can also see where automation might lose important context, create checking work or introduce risk.

Meaningful involvement should go beyond asking whether employees support AI. More useful questions include:

• Which parts of the work could reasonably be assisted?
• Which decisions require human judgement or discretion?
• Where could important context be lost?
• What kinds of error would have serious consequences?
• How should AI-generated work be checked?
• What would make the technology genuinely useful in the existing workflow?

This also helps employers make a more careful distinction between tasks AI can support and work that should retain a human at its centre.

Build capability, not just access

Giving employees an AI licence is not the same as enabling them to use it safely and effectively. Training should address not only how the tool operates, but also the judgement required around it.

Employees need to understand acceptable and prohibited uses, privacy and confidentiality requirements, how to assess reliability, when independent checking is required and who remains accountable for the final work. Managers need additional capability because employees will look to them for practical guidance and for signals about whether experimentation, questions and challenge are genuinely safe.

Protect psychological safety

Employees need to be able to admit that they do not understand a tool, challenge an unreliable output and report a mistake without being labelled resistant or incapable.

Employers should also avoid creating an artificial divide between enthusiastic early adopters and employees who need more time or support. Adoption speed is not necessarily a measure of competence, commitment or future value. A cautious employee may be identifying risks that a confident early adopter has overlooked.

Pay attention to what happens to the time saved
One of the most important questions may be what the organisation does with any capacity AI releases.

If AI saves time but that time is immediately absorbed by greater volume, tighter targets or reduced resourcing, employees may experience the technology as work intensification rather than support. Productivity may rise while wellbeing deteriorates.

Employers should therefore measure more than output. They should ask whether the tool has reduced frustration, improved quality, supported a sustainable pace of work, created new checking or correction tasks, or changed performance expectations without those changes being openly discussed.

The enduring people principles

AI adoption should be treated as a people change, not simply a technology rollout. The organisations most likely to gain sustainable value will not necessarily be those that adopt the fastest. They will be those that explain the purpose clearly, involve employees in redesigning the work, build capability, preserve psychological safety and remain honest about what the technology may mean over time.

AI changes the tools available to us. It does not change what people need in order to navigate change well.

Source

Article referenced: How HP's ANZ people leader is using AI to ease workplace stress, Human Resources Director, 28 July 2026

In the next article in this series, Are You Using AI to Think Better - or to Think Less?, we shift the focus from how organisations manage AI adoption to how individuals use it; exploring whether AI is helping us think better, or simply doing more of the thinking for us.