
What changed in NSW on 1 July - and what it means for the driver nobody is naming
A document that had been guidance for five years became a benchmark. Most organisations have not changed anything they do.
On 1 July 2026, section 26A of the Work Health and Safety Act 2011 (NSW) commenced, inserted by the Industrial Relations and Other Legislation Amendment (Workplace Protections) Act 2025.
It is short. A person conducting a business or undertaking must either comply with an approved code of practice, or manage the risk in a way that provides an equivalent or higher standard of work health and safety.
Read quickly, that sounds like what was already true. It is not.
What codes of practice used to be, and what they are now
Codes of practice have always been practical guidance on how to meet a duty. They could be relied on in proceedings as evidence of what was reasonably practicable in the circumstances, and an organisation was free to take a different route provided the route met the standard.
The difference now is the word demonstrate. Under section 26A, an alternative approach has to be shown to deliver an equivalent or higher standard - which means the comparison has to exist somewhere other than in the head of the person who designed the system.
One of the codes caught by this is Managing psychosocial hazards at work. It moved, in a single day, from a document that described good practice to a document that functions as the benchmark you are measured against.
The context it landed in
This did not arrive alone.
The Work Health and Safety Regulation 2025 (NSW) commenced on 22 August 2025 and requires psychosocial risk to be managed through the hierarchy of control measures. Higher-order controls first. Training and individual coping strategies are not prohibited, but they sit near the bottom of the hierarchy, and an organisation relying on them is expected to be able to say what it considered above them.
Psychosocial risk is one of four regulatory priorities in SafeWork NSW’s Regulatory Statement 2026-27, alongside falls from heights, hazardous substances, and mobile plant and machinery. The Statement names employer compliance with the enforceable codes of practice as a focus across all four.
And the enforcement is resourced. In March 2026 SafeWork NSW deployed twenty dedicated psychosocial safety inspectors as part of its largest inspectorate expansion, and has since established a Psychosocial Advisory Service, supported by a $127.7 million commitment from the NSW Government. Over the preceding twelve months it received more than 2,200 requests for service and was notified of more than 190 incidents relating to psychosocial hazards.
None of that is a prediction. It is the published position of the regulator, and it is unusual for a regulator to say this much this plainly.
What it changes in practice
Three things, and this is my reading of the implications rather than an interpretation of the law.
Documentation stops being a project and becomes a standing state. The question is no longer whether you did the work, but whether you can produce the evidence of it on the day you are asked.
A training-led response now needs a written record of what sat above it. If the primary control for a psychosocial risk is a session, an app or a policy, the hierarchy asks what structural change was considered first and why it was not reasonably practicable. That reasoning has to exist in writing before someone asks for it, not after.
Consultation has to be evidenced as consultation. An annual engagement survey is a measurement instrument. It is not, on its own, a record of workers being consulted about the risks and controls that affect them.
Where financial pressure sits in this
Precisely here, and no further.
Financial pressure is not a listed psychosocial hazard under the Code, and nothing in this piece argues that it should be. It operates as a driver and an amplifier of hazards that are listed - job demands, low job control, poor support - which is a smaller claim and a more defensible one.
But comply-or-justify sharpens why that matters commercially.
Equivalence is judged on the standard of health and safety actually delivered, not on the tidiness of the paperwork describing it. An organisation can hold a complete hazard register, map each hazard to a control, review on schedule, and still be operating controls that are quietly performing below what the documentation says - because something is running underneath them that appears nowhere in the register, under any name.
A support mechanism nobody uses is a control on paper. A workload adjustment offered to someone whose real constraint is arriving from outside the workplace is a control aimed slightly off-target. Neither shows up as a failure until something else does.
The Code did not create that problem. It just raised the standard of proof on controls, which makes any driver you cannot see more expensive than it was in June.
This is a summary of a regulatory change, not legal advice. Organisations should take their own advice on how section 26A applies to them. Sources: Work Health and Safety Act 2011 (NSW) s 26A; Work Health and Safety Regulation 2025 (NSW); SafeWork NSW, Regulatory Statement 2026-27; SafeWork NSW media release, 1 July 2026.
If you want the evidence base underneath this
The Hidden Patterns Beneath Performance is my research brief on financial pressure as a driver of psychosocial risk. Every figure sourced, dated and scoped, including what the evidence does not yet support.