The Future of Workplace Safety: AI, Emerging Risks and Smarter Risk Management
Author: Alex Tillman, Principal Consultant
From Reactive to Predictive: AI’s Role in the Future of Workplace Safety
Artificial intelligence (AI) is rapidly reshaping workplace health and safety (WHS), creating both transformative opportunities and complex emerging risks. For Australian organisations, this shift is no longer theoretical—regulators are explicitly bringing AI into scope of WHS obligations, making it a board‑level issue for 2026 and beyond.
A new class of WHS hazards
Australia’s Work Health and Safety Strategy 2023–2033 identifies AI, automation and new technologies as key emerging challenges which aligns to what we have been seeing with our clients. While these technologies can reduce exposure to physical hazards, they may simultaneously introduce new psychosocial risks, including increased job insecurity, cognitive load, complex decision-making, and changing role clarity.
Globally, research reinforces this shift: AI creates “novel cognitive, psychosocial, organisational and ethical challenges” arising from human–machine interaction and algorithmic decision-making. At the same time, the International Labour Organisation notes that AI-enabled systems—from robotics to predictive analytics—can reduce hazardous exposures and improve safety outcomes, while also introducing risks such as over-reliance on automation, surveillance-related stress, and ergonomic impacts.
Regulatory momentum in Australia
Australian regulators are moving quickly to respond. In February 2026, NSW passed the WHS Digital Work Systems Act 2026, explicitly extending WHS duties to AI and algorithmic systems. Under these reforms, organisations must ensure—so far as reasonably practicable—that AI systems do not create health and safety risks, including risks arising from how work is allocated or monitored.
Critically, the legislation highlights specific risk categories, including:
Excessive workloads generated by AI systems
Intrusive monitoring or surveillance
Unreasonable performance metrics
Algorithmic bias or discriminatory decision-making
This marks a significant shift: technology itself is now clearly recognised as a WHS hazard, and not just a productivity tool. Similar regulatory tightening is expected nationally, with WHS laws already requiring businesses to manage both physical and psychosocial risks arising from AI use. NSW is leading with explicit regulation, while other states are regulating AI “by stealth” through psychosocial and general WHS duties. As AI adoption accelerates, organisations are increasingly adopting more structured approaches to governance and oversight, with a strong focus on accountability, transparency, and effective risk management.
The opportunity: from reactive to predictive safety
Despite these risks, AI presents a step-change opportunity for WHS. Traditional safety systems tend to be reactive—relying on incident reports and lag indicators. AI enables a more proactive model, using real-time data, predictive analytics, and monitoring technologies to identify hazards before incidents occur.
Common applications include:
Predictive modelling of incidents and injuries
Computer vision for hazard detection
Wearables and sensors for real-time risk monitorin
Automated analysis of safety data and near misses
These tools support a shift toward anticipatory risk management, improving both safety outcomes and operational efficiency.
The emerging risk: psychosocial and governance challenges
However, the most significant emerging risks are not physical—they are psychosocial. AI-driven systems that allocate tasks, monitor performance, or optimise productivity can unintentionally increase:
Work intensity and fatigue
Perceived surveillance and loss of autonomy
Job insecurity and distrust
Organisations need to actively assess and control psychosocial hazards—not just document policies. In most cases, the risk does not come from the technology itself, but from how the organisation designs and uses it.
What this means for organisations
The key implication is clear: AI must be embedded into existing WHS risk management frameworks. Organisations should:
Treat AI systems as hazards requiring formal risk assessment
Evaluate both physical and psychosocial impacts of workplace changes AI brings
Ensure transparency, consultation, and worker engagement
Implement governance and oversight for AI decision-making
We’ve seen some clients embrace AI and do all of the above effectively, but there is a need to stay agile alongside the fast pace of technological change.
AI represents one of the most significant shifts in workplace safety in decades. It offers the potential to dramatically improve hazard identification and prevention—but also introduces new risks that are less visible and more complex. Those that proactively integrate AI into their WHS systems will not only meet compliance expectations, but also unlock safer, more resilient, and more productive workplaces.
More Information:
AIHA: ILO Report Highlights Risks and Benefits of Artificial Intelligence
