Executive Summary

Every month, Australian boards receive a WHS or people risk report. It includes workers' compensation claim numbers, injury frequency rates, return-to-work statistics, and perhaps an employee engagement survey score. Directors review the figures, ask questions, satisfy themselves that management is across the issues, and move on to the next agenda item.

The problem is not the questions directors ask. The problem is that the data they are reviewing is structurally incomplete - and the gap is most acute, and most consequential, in the one category of risk that is growing fastest.

Mental health conditions now account for 12% of all serious workers' compensation claims nationally and have increased by 161% over the past decade, according to Safe Work Australia. In NSW alone, psychological injury claims now make up 12% of total workers' compensation claims but consume 38% of total cost, with the average claim cost rising from $146,000 in 2019-20 to $288,542 in 2024-25, according to NSW Treasurer Daniel Mookhey's March 2025 ministerial statement. Claims have doubled in six years. The workers' compensation system is under actuarial stress. Premiums are increasing. And this is the data that reaches boards.

What does not reach boards is the far larger volume of psychosocial harm that never becomes a compensation claim: the workers experiencing burnout who resign rather than report, the harassment that is witnessed but not escalated, the team cultures corroding under unmanaged workload, the manager whose behaviour is an open secret but has not yet generated a formal complaint. Workers' compensation data captures the injuries serious enough that workers, after considering stigma, employment consequences, complexity, and the claims process, decide to lodge. It is the visible fraction of a much larger problem.

For directors, this creates a governance paradox. The duty of care requires proactive risk management - identifying and controlling hazards before workers are harmed. The Australian Institute of Company Directors (AICD), in its 2024 primer for boards on governing WHS psychosocial risks developed in partnership with King & Wood Mallesons, is unambiguous: directors must take reasonable steps to understand psychosocial risks and confirm that appropriate resources and processes are in place to manage them. That duty cannot be discharged on the basis of data that systematically fails to capture the problem it is meant to surface.

This article examines the structural gap between the psychosocial harm that occurs in Australian workplaces and the data that reaches board dashboards, why that gap is legally consequential, what good board-level reporting on psychosocial risk should look like, and why the reporting infrastructure organisations build today will determine not only their regulatory exposure but their insurance costs, premium trajectories, and long-term financial resilience.

The Scale of the Problem That Workers' Compensation Data Does Not Fully Capture

The workers' compensation claims data that boards routinely review is the most consequential available measure of psychosocial harm in Australian workplaces. It is also, by definition, a lagging measure of harm that has already occurred, has been serious enough to generate a formal claim, and has survived the significant barriers that prevent many injured workers from lodging.

The national data is stark. According to Safe Work Australia's Key Work Health and Safety Statistics, mental health conditions accounted for 17,600 serious workers' compensation claims in 2023-24, representing 12% of all serious claims and a 161% increase over the preceding decade. Median time off work for mental health claims is 35.7 weeks - more than five times the median for physical injuries. Median compensation paid per mental health claim is $67,400, compared to the overall median of $16,300 across all claim types.

In New South Wales, the state with the most detailed publicly available scheme data, the picture is even more striking. In a March 2025 ministerial statement, NSW Treasurer Daniel Mookhey told parliament that psychological injury claims, while 12% of total claims by volume, account for 38% of the total cost of the workers' compensation scheme. The average cost of a psychological injury claim in NSW has risen from $146,000 in 2019-20 to $288,542 in 2024-25. Claims have doubled in six years. At current trajectories, iCare actuaries project that without structural change, the scheme's deficit could blow out to $6 billion.

WorkCover WA data similarly reveals that claims exceeding $500,000 - classified as large losses - have nearly quadrupled between 2014 and 2023, with their share of total scheme costs expanding by 233% over nine years.

These figures represent the injuries that have cleared the threshold: a worker who has experienced psychological harm serious enough to seek medical attention, who has been assessed as work-related, who has navigated the claims process, and whose claim has been accepted. Safe Work Australia's own analysis of the National Dataset for Compensation-based Statistics explicitly notes that it "captures only injuries and illnesses that result in accepted workers' compensation claims rather than the full universe of workplace harm."

The structural analysis firm Safetysure, in its 2025 review of Safe Work Australia data, reached the same conclusion: the 17,600 serious mental health claims in workers' compensation statistics may represent an uncertain fraction of the total burden of work-related mental health conditions experienced by Australian workers, with mental health conditions potentially having higher rates of non-claiming due to stigma, concerns about employment consequences, difficulty establishing work-relatedness, or uncertainty about eligibility.

The harm the board sees in its WHS report is not the harm occurring in the organisation. It is the harm that survived every barrier between experience and claim.

The Governance Gap: What the Due Diligence Obligation Actually Requires

The WHS officer duty under Australia's model Work Health and Safety Act is not a duty to review claims data. It is a duty to exercise due diligence to ensure the organisation complies with its WHS obligations - which includes taking reasonable steps to understand psychosocial hazards in the workplace and confirming that appropriate resources and processes are in place to manage them.

The AICD's 2024 Governing WHS Psychosocial Risks: A Primer for Boards, developed in partnership with King & Wood Mallesons, sets out what this duty requires in practice. It is direct: directors must take "reasonable steps to understand psychosocial risks in the workplace and confirming that appropriate resources and processes are in place to manage them."

The primer identifies specific red flags that should concern directors, including:

• The absence of a robust safety governance plan

  
  

• Lack of board reporting on psychosocial risks

  
  

• Failure to act on identified risks

  
  

• Insufficient training on psychological health and safety

  
  

The primer further specifies that directors exercising due diligence should be able to answer the following questions of their management teams:

• Has the organisation carried out a risk assessment of psychosocial hazards in consultation with the workforce, using good data analytics and sound risk analysis techniques?

  
  

• Which of the fourteen psychosocial hazards are most critical to the organisation?

  
  

• Has accountability been allocated to implement, measure and monitor proactive controls to minimise risk?

  
  

• Can management provide the board with reporting on high-risk controls, including lead performance indicator reporting?

  
  

The critical phrase is "lead performance indicator reporting." Boards that are receiving only lagging indicators - claims, incidents, formal complaints - are not receiving what the AICD describes as sufficient board-level reporting on psychosocial risk. They are receiving evidence of harm that has already occurred. They are not receiving the forward-looking data that allows them to exercise the proactive oversight the law requires.

Ashurst Risk Advisory Partner Tony Morris, writing in the AICD's Company Director magazine, put the board's information obligation in direct terms: in large organisations, a common and effective way for boards to exercise due diligence is through "well organised, insightful and action-oriented reporting; providing both lead and lag indicators to measure the effectiveness of the risk environment."

A board that is receiving only lag indicators is not exercising due diligence. It is reviewing the casualties after the fact.

The Data Architecture Problem: Why Current Reporting Fails Boards

The structural gap between psychosocial harm and board-visible data is not primarily a matter of organisational negligence. It reflects the architecture of the reporting systems that most organisations have inherited.

Traditional WHS reporting was designed around physical hazards - incidents, injuries, near-misses. The reporting architecture: wait for an event, record it, report it upward. This architecture works reasonably well for physical harm because the signal - a broken bone, a chemical exposure, a fall from height - is discrete, observable, and typically reported with relative speed.

Psychosocial harm does not work this way. It is cumulative, gradual, often invisible to managers and peers, frequently attributed by affected workers to personal circumstances rather than work conditions, and carries specific barriers to reporting that physical harm does not. The result is that the incident-based reporting architecture inherited from physical WHS systematically fails to surface psychosocial risk at the speed and completeness required for effective management.

Consider what happens to a worker experiencing harm from chronic work pressure, or persistent harassment, or an unsupported return from leave. The worker is unlikely to lodge an incident report. They may speak to a trusted colleague, say nothing to their manager, reduce engagement, increase sick leave, and eventually either resign or reach a point at which the harm is severe enough to prompt a medical consultation, a formal complaint, or a workers' compensation claim. By the time the harm reaches the board in any data form, it has been months or years in development.

The AICD's primer is explicit about this: "Psychosocial incidents are as damaging as physical harm - but they're harder to spot. Reporting is less mature and incidents can escalate fast if not mitigated early. They might only become apparent after harm has occurred." Helen Rowell MAICD, a director at Mental Health Australia, noted at the AICD's primer launch event that there is no single measure that can tell an organisation whether it is doing well or poorly on psychosocial hazard risk: "It's actually about trying to make sure you've got a good picture of what's happening across the organisation using a range of statistics - then seeing where that leads you to dig deeper and understand more."

This insight points to the solution. Boards need data from multiple sources, capturing signals at different stages of the harm trajectory, not just at the end point. Workers' compensation claims tell you about harm that has already been serious for a long time. What boards need to be able to see is the earlier signal: the pattern of concerns being raised confidentially, the team-level reporting trends, the early indicators that a psychosocial hazard is operating at intensity in a particular part of the organisation before it reaches claim-level severity.

What Good Board-Level Psychosocial Reporting Looks Like

The Australian WHS Strategy 2023-2033 sets a clear national objective: to build the capability of PCBUs, WHS regulators, and workers to ensure compliance with the duty to manage psychosocial hazards. SafeWork NSW's Psychological Health and Safety Strategy 2024-2026 signals that inspectors will complete Psychosocial WHS Checks when visiting larger organisations, and that businesses should expect additional planned compliance visits and revisits for non-compliance. The direction is clear: boards that cannot demonstrate evidence-based, proactive psychosocial risk oversight are increasingly exposed.

Effective board-level psychosocial reporting integrates four categories of information that together provide a genuine picture of risk - not just confirmed harm.

1. Lead Indicators: Early Risk Signals

Lead indicators capture signals before they become incidents. In the psychosocial context, these include: the volume and nature of concerns being raised through confidential reporting channels; patterns of early intervention requests; team-level reporting rates relative to workforce size; manager-level feedback data; and the output of psychosocial risk assessments conducted in consultation with workers. An organisation whose lead indicator data shows a particular team reporting elevated workload concerns, with declining confidence in management support and increasing absence rates, is seeing a psychosocial risk signal that is weeks or months ahead of a potential claim.

2. Lag Indicators: Confirmed Harm Data With Context

Lag indicators remain essential - workers' compensation claims, incident reports, formal complaints, return-to-work outcomes. But the AICD's due diligence framework requires that lag indicator data be contextualised against the lead indicators, not reviewed in isolation. A board that sees a reduction in formal psychological injury claims this quarter cannot conclude that psychosocial risk has reduced, without also examining whether lead indicators - confidential reporting rates, early intervention requests, engagement data - have moved in the same direction. A reduction in claims accompanied by a reduction in reporting confidence is evidence of a deteriorating reporting culture, not improving safety.

3. Root Cause Analysis and Control Effectiveness

The model WHS Regulations require PCBUs to implement control measures and review their effectiveness. Board reporting should include not just incident and claim data, but analysis of what psychosocial hazards the controls are targeting, what evidence exists that they are working, and what the mechanism of harm was for completed claims. The breakdown of mental health claim causes reported in Safe Work Australia data - workplace harassment and bullying (33.2% of mental stress claims), work pressure (24.2%), and exposure to workplace violence (15.7%) - represents the aggregate national picture. Boards should be receiving the equivalent analysis for their own organisation, with evidence of the specific controls targeting each category.

4. Jurisdictional and Sector Benchmarking

With Victoria's new Occupational Health and Safety (Psychological Health) Regulations 2025 completing the national framework from December 2025 - meaning every Australian jurisdiction now requires employers to explicitly identify, assess, and control psychosocial hazards - boards of organisations operating across multiple states face a layered compliance landscape. Board reporting should include a jurisdictional compliance map showing the specific regulatory requirements in each state where the organisation operates, and confirmation that the organisation's psychosocial risk management framework meets the requirements of the most demanding applicable regime.

The Financial Case: Why Psychosocial Data Quality Is a Balance Sheet Issue

Directors who approach psychosocial risk primarily as a regulatory compliance matter are missing its financial dimension, which has become sufficiently material that it is attracting actuarial analysis, scheme restructuring, and explicit parliamentary attention.

The NSW workers' compensation scheme is instructive. The Workers Insurance Scheme recorded a net loss of $1.73 billion in 2024-25. The scheme's net outstanding claims liability has climbed to $25.6 billion. Actuaries explicitly factor the incidence of future psychological claims into their liability projections and warn that these claims "have significantly longer durations and higher costs associated with them."

For individual employers, this translates directly into premium exposure. Australian organisations with poor psychosocial safety records - whether demonstrated by higher claims rates, longer durations, or regulator findings - pay more. The iCare framework in NSW, like premium structures across other jurisdictions, links an employer's workers' compensation costs to their own claims performance. The HRD Australia analysis of iCare's 2024-25 report is pointed: investments in training, early intervention, confidential reporting channels, and evidence-based mental health programs "now double as a form of balance-sheet protection."

Nationally, claims involving more than 13 weeks off work - the category that mental health claims disproportionately fall into - accounted for 21.9% of total claims in 2022-23 but 74.8% of total compensation payments, representing $5.4 billion in compensation costs (Safe Work Australia). For any employer with a meaningful workforce, preventing a single long-duration psychological injury claim from progressing through to a multi-year compensation engagement is worth far more than the cost of the reporting infrastructure and early intervention systems that might have prevented it.

The Committee for Economic Development of Australia (CEDA) projected in 2021 that mental health claims would double to over 20,000 annually and reach $1 billion in cost by 2030. Based on 2024-25 data from CEDA and InCheq, that threshold has been reached five years early. Mental injury compensation claims crossed $1 billion in 2024-25 - five years ahead of the projected milestone. For directors, the trend line is a financial risk signal, not just a health and safety one.

The Personal Liability Dimension: What "Due Diligence" Means for Individual Directors

The WHS officer duty is personal and non-delegable. Under the model WHS Act, an officer who fails to exercise due diligence to ensure the PCBU complies with its WHS duties commits an offence. The penalties are substantial: under Category 2 offences, corporations face fines up to $3.6 million and individuals up to $360,000; under Category 1 (reckless conduct exposing a person to serious harm), corporations face up to $3.45 million and individuals up to 5 years imprisonment. Industrial manslaughter provisions, now in force in every jurisdiction except South Australia and the Northern Territory, carry up to 25 years imprisonment for individuals in some states.

The AICD notes that in nearly all Australian jurisdictions, there is a positive obligation on directors to exercise due diligence in relation to WHS - and that the penalties extend to possible imprisonment and substantial fines. The organisation has also noted a shift in regulatory approach: non-executive directors are increasingly included in prosecutorial action, not just executive directors, with regulators examining what steps the board as a whole has taken rather than focusing solely on individual executives.

What constitutes evidence of due diligence? The AICD's primer provides a framework. For psychosocial risk specifically, evidence of due diligence includes:

• Demonstrated understanding of the psychosocial hazards specific to the organisation's operations

  
  

• Evidence of proactive risk assessment conducted in consultation with the workforce

  
  

• Board reporting that includes both lead and lag indicators of psychosocial risk

  
  

• Evidence that controls have been implemented and their effectiveness reviewed

  
  

• Documentation of the board's active engagement with management on psychosocial risk management

  
  

A director who can produce evidence that the board was regularly receiving comprehensive psychosocial risk reporting, was actively questioning management on the lead indicators, and was satisfied that controls were in place and being reviewed is in a materially different position to a director who reviewed claims data and assumed the situation was under control.

Crucially, the duty to verify is personal. Herbert Smith Freehills managing partner Steve Bell, in AICD commentary, put it directly: "Directors must be satisfied that there is a plan of action to assess and control the risks." Satisfaction requires evidence. And evidence of proactive psychosocial risk management is only possible if the reporting infrastructure exists to generate it.

The Practical Governance Questions Every Director Should Be Asking Now

Against this backdrop, the following questions represent the minimum threshold for a director exercising due diligence on psychosocial risk in 2026:

1. What data does our board actually receive on psychosocial risk, and is it lead or lag?

If the psychosocial data in board reporting consists exclusively of workers' compensation claims, incident reports, and annual engagement survey scores, the board is receiving lag indicators only. The AICD and WHS regulators are explicit that this is insufficient. Directors should request a review of the reporting framework with specific focus on what lead indicators are being captured, how frequently, and at what level of granularity.

2. Do we know what percentage of our workforce's psychosocial harm is captured by our current reporting systems?

The workers' compensation data that reaches boards represents confirmed harm that has cleared significant barriers to reporting. The actual prevalence of psychosocial harm in Australian workplaces substantially exceeds what claims data captures. Directors should ask management: what is our estimate of the gap between experienced harm and reported harm in our organisation? What evidence do we have for that estimate? What are we doing to reduce the gap?

3. Are our confidential reporting channels trusted enough to be used, and do we have evidence of this?

A confidential reporting channel that has generated zero psychosocial risk reports from a workforce of 300 people in the past 12 months is not evidence of a healthy workplace. It is evidence of a reporting failure. The AICD Life Street CEO Paul Flanagan has noted that from an individual's perspective, a workers' compensation claim is often the last resort. The systems that exist between experience and claim - informal support, early conversation, confidential reporting, EAP, management escalation - determine whether harm is caught early or reaches the point of formal claim. Directors should be asking: are our systems genuinely accessible and trusted, and how do we know?

4. Are we benchmarking our psychosocial risk profile against sector and national data?

National data on the causes of mental health claims - workplace harassment and bullying (33.2%), work pressure (24.2%), and workplace violence (15.7%) as identified in Safe Work Australia analysis - gives boards a framework for calibrating their organisation's risk profile. Does the organisation's internal data show a similar distribution, or are certain categories disproportionately represented? What does that tell the board about where the priority controls should be?

5. Is our psychosocial risk management framework being reviewed, or is it static?

The model WHS Regulations require that control measures be reviewed and revised. A psychosocial risk management framework that was established in response to regulatory change and has not been substantively reviewed since is unlikely to meet the standard of ongoing due diligence. Directors should be receiving evidence of review cycles: what controls were tested, what the evidence showed, and what changes were made as a result.

Key Takeaways

• Mental health conditions account for 12% of all serious workers' compensation claims nationally, with a 161% increase over the past decade. In NSW, they account for 38% of total scheme cost despite being 12% of claims by volume. The average cost per claim in NSW has risen to $288,542. Mental injury compensation nationally crossed $1 billion in 2024-25, five years ahead of projections.

  
  

• Workers' compensation data represents only those injuries serious enough to generate accepted claims after clearing significant barriers. Safe Work Australia explicitly acknowledges the National Dataset captures only the injuries that result in accepted claims - not the full universe of workplace harm. The data directors receive on psychosocial risk is structurally incomplete.

  
  

• The AICD's 2024 Governing WHS Psychosocial Risks: A Primer for Boards, developed with King & Wood Mallesons, states that directors must take reasonable steps to understand psychosocial risks and confirm that appropriate resources and processes are in place. This duty requires both lead and lag indicators - not claims data alone.

  
  

• Directors have personal, non-delegable due diligence obligations under WHS law. The penalties for failure include substantial fines and potential imprisonment. Regulators are increasingly including non-executive directors in prosecutorial actions. Evidence of due diligence requires evidence of proactive oversight, not just review of harm that has already occurred.

  
  

• SafeWork NSW's Psychological Health and Safety Strategy 2024-2026 signals active compliance inspections, Psychosocial WHS Checks at larger organisations, and naming of repeat offenders. Victoria's new Occupational Health and Safety (Psychological Health) Regulations 2025 completing the national framework from December 2025 means every jurisdiction now requires explicit psychosocial hazard identification, assessment, and control.

  
  

• Confidential reporting channels that generate no data are evidence of a reporting failure, not a safe workplace. Boards need evidence that their reporting infrastructure is trusted, accessible, and generating the early-signal data that allows proactive intervention before harm escalates to claim level.

  
  

• Salus provides the lead indicator reporting infrastructure that Australian boards need to discharge their psychosocial due diligence obligations: a confidential, 24/7 reporting platform that captures psychosocial risk signals across the workforce, generates longitudinal trend data visible to the board, and provides the evidence base that transforms a governance aspiration into a demonstrable, auditable safety practice.

  
  

If the psychosocial data reaching your board consists primarily of workers' compensation claims and annual survey scores, you are exercising governance on structurally incomplete information in the fastest-growing and highest-cost risk category in Australian WHS. Safe Work Tech's Salus platform provides the lead indicator reporting infrastructure boards need - confidential, real-time, longitudinal, and designed to surface psychosocial risk before it reaches compensation claim severity. Contact Safe Work Tech to understand how Salus can close the gap between what your board is seeing and what is actually happening in your organisation.