For a long time, psychological safety was treated as a culture issue. Important, yes, but often soft around the edges. Something for HR to support, managers to encourage, and employees to speak up about if things became too much.
That era is over.
Across Australia, employers are expected to identify psychosocial hazards, assess the risks, implement effective controls, and review whether those controls are actually working. Safe Work Australia’s Model Code of Practice makes it clear that PCBUs must eliminate or minimise psychosocial risks so far as is reasonably practicable. In Victoria, the Occupational Health and Safety (Psychological Health) Regulations 2025 have now commenced, creating specific duties for employers to identify psychosocial hazards, control associated risks and review those controls.
Psychosocial hazards are factors in the design or management of work that can increase the risk of work-related stress and lead to psychological or physical harm. WorkSafe Victoria lists examples such as poor supervisor support, high job demands, low job control, poor workplace relationships, fatigue, violence, traumatic events and bullying.
These hazards rarely appear in isolation. A team may be dealing with excessive workload, unclear priorities, poor change communication and a difficult manager at the same time. Each factor might look manageable on its own. Combined, they can become a serious risk.
That is where organisations get caught. They look for a single incident, a single complaint or a single “problem employee”. Regulators are looking for something broader: whether the organisation had a system to identify foreseeable risks, consult with workers, implement controls and test whether those controls reduced harm.
Many organisations still respond to psychosocial risk by updating a policy, running a training session or reminding employees about the EAP. Those things may help, but they are not enough.
The current regulatory direction is toward higher-order controls: changing work design, management systems, resourcing, role clarity, supervision, rostering, escalation pathways and the way work is actually done. SafeWork NSW’s guidance, for example, says employers should start with higher-order controls such as redesigning how work is completed before relying on lower-order controls such as training or coping strategies.
That should make leaders pause. If your main control for excessive workload is resilience training, you may not have controlled the risk. If your response to poor behaviour is a once-a-year code of conduct refresher, you may not have controlled the risk. If your only signal is complaints, you may be finding out too late.
Leaders do not need to create a perfect workplace. They do need to do what a prudent employer would do, based on the likelihood of harm, the severity of potential harm, what is known about the hazard, what controls are available, and whether those controls are suitable.
In practice, that means asking:
If the answer to those questions is vague, the risk is not just cultural. It is legal, operational and reputational.
The real test is not whether your organisation has a psychological safety policy. It is whether leaders can prove they know where the risks are, what they are doing about them, and whether those actions are reducing harm.
That requires evidence: a psychosocial risk register, consultation records, documented control decisions, action plans, review dates, trend data and leadership reporting. Your board and executive team should be seeing leading indicators, not just lagging outcomes after harm has already occurred.
Mentally safe work is not about removing every pressure from the workplace. Pressure, deadlines and difficult work will always exist. The obligation is to design and manage work so those pressures do not become foreseeable, preventable harm.
For leaders, this is the moment to get ahead of the risk. Because when psychosocial hazards are ignored, they do not stay invisible. They show up in people data, performance issues, claims, investigations and culture. By then, the cost is already much higher than prevention.