AI adoption is not landing evenly, and that’s a psychosocial risk

AI’s potential to improve efficiency and enhance your employees’ experience of work is huge, but the organisations that will get AI right aren’t the ones moving fastest, writes Ingrid Jenkins.

When organisations talk about AI at work, the conversation usually centres on productivity. How much time it saves. How many tasks it can automate. How it enables teams to work faster.

And while these matter, they’re only part of the story. The question worth asking alongside them is how our people are actually keeping up.

AI rollouts can look very different depending on your role and on where you sit in an organisation. For some employees, it’s energising. This is a tool that removes friction and creates space for more meaningful work. But for others, it’s quietly destabilising. Expectations shift without explanation, and colleagues who’ve adopted AI faster seem to be pulling ahead.

This isn’t a culture problem. It’s a design problem. And in Australia, it’s increasingly a legal one.

The uneven reality of AI adoption

AI capability isn’t landing evenly across organisations, and that unevenness is a psychosocial hazard in its own right. People navigating additional barriers, whether that’s lower digital confidence, language differences, or role types that haven’t been prioritised in rollout planning, are often the least supported through the transition.

According to Sonder support cases, job demands, poor support, and lack of role clarity are the three most common psychosocial hazards experienced by workers. AI, without careful implementation, intensifies all three simultaneously. When people don’t know where AI fits into their role, who’s accountable when it gets something wrong, or whether they’re expected to work at the pace the machine sets, that ambiguity doesn’t just create confusion. It creates risk.

When your approach to AI looks the same for everyone, it’s probably not working for everyone. Customising how you build AI capability across your workforce isn’t just a nice-to-have. Under Australia’s work health and safety legislation, employers have a duty to identify and manage psychosocial hazards. An uneven AI rollout creates identifiable hazards around role clarity, job control, and support. If your last psychosocial risk assessment predates your AI adoption, it’s out of date.

What good looks like

Organisations getting this right aren’t treating AI as a tool dropped into existing workflows. They’re treating it as a structural change to the work environment, one that requires the same discipline as any other significant change to how work gets done.

That means embedding AI into role design, not layering it on top of existing demands. It means building role-specific training rather than assuming a company-wide program will land the same way for a frontline worker as it does for a knowledge worker. And it means being explicit, in writing, about where AI fits into a process and where it doesn’t.

When accountability is unclear, the hazards of poor role clarity and low job control follow quickly. Employees are left asking, “Am I responsible if the AI gets it wrong?” “What part of this work is actually mine now?” Those questions, left unanswered, don’t stay theoretical. They show up as disengagement, hesitation, and the kind of quiet withdrawal that managers often misread as resistance.

Reframe resistance as a signal

When employees struggle with AI tools, it tends to be labelled as resistance to change. In most cases, it’s something else entirely, a lack of support, clarity, or confidence that hasn’t been addressed. Before concluding you have an adoption problem, ask what your people are actually missing.

AI’s potential to improve efficiency and enhance your employees’ experience of work is huge, but the organisations that will get AI right aren’t the ones moving fastest. They’re the ones moving with enough intention to bring everyone with them and recognising that the cost of leaving people behind isn’t just a productivity problem. It’s a safety one.

Ingrid Jenkins is the chief people officer at Sonder.

Want to see more stories from trusted news sources?
Make HR Leader a preferred news source on Google.