AI at Work: What SafeWork Australia's New Guidance Means for Your Business
If your business has started using AI tools, automation, wearables, or new software in the last year or two, there's a good chance nobody has stopped to ask: is this actually safe for our people?
SafeWork Australia has just released dedicated guidance on AI and digital technologies in the workplace, and it confirms something we've been telling our clients for a while now: these tools aren't just an IT decision or a productivity upgrade. They come with work health and safety (WHS) obligations, the same as any machine, chemical, or work process on your site.
What counts as "AI and digital technology"?
The definition is broader than most business owners expect. It's not just chatbots and generative AI. It includes:
Computers, mobile devices, and workplace software or apps
Security and safety systems
Wearable devices
Advanced robotics and automation, including cobots (collaborative robots)
Machine learning tools and predictive software used to schedule, monitor, or direct work
If your business uses rostering software, tracking apps, automated machinery, or AI-assisted decision tools, this guidance applies to you.
The core message: manage it like any other risk
SafeWork Australia is clear that when you introduce or change AI or digital technology in your workplace, you must manage the associated risks the same way you would manage any other WHS risk. That means:
Identify the hazards. What could this technology do, physically or psychologically, to the people using it or working alongside it?
Assess the risk. How likely is harm, and how serious could it be?
Control the risk. Eliminate it if you can. If you can't, minimise it so far as is reasonably practicable.
Consult your workers. This isn't optional. Workers and health and safety representatives need a genuine opportunity to raise concerns and contribute to decisions before and after the technology is introduced.
Importantly, this applies even to changes you didn't initiate yourselves, such as an automatic software update pushed by a vendor, or a "routine" IT change. If it changes how work is done, it needs a risk assessment.
The risks businesses are missing
Some of the risks SafeWork Australia highlights are physical: automated gates malfunctioning, or a cobot moving unexpectedly and colliding with a worker. These are the ones businesses tend to think about first.
But a lot of the risk is psychosocial, and far less visible:
Work intensification and fatigue. Automating the routine parts of a job can leave workers with a heavier load of complex, high-concentration tasks, increasing fatigue risk.
Excessive surveillance. Monitoring tools can create pressure and erode trust if introduced without transparency or consultation.
Unfair or discriminatory decision making. Algorithms used for scheduling, performance monitoring, or task allocation can produce outcomes that disadvantage certain workers, often without anyone intending it.
Low job demands. At the other end, automation can strip roles of enough meaningful work, which is its own psychosocial risk.
It's not all risk, there's opportunity too
Used well, AI and digital technology can genuinely improve safety outcomes: think adjustable settings that prioritise health and safety, or accessibility features that reduce risk exposure for workers who need them. The point isn't to avoid these tools. It's to introduce them deliberately, with the same rigour you'd apply to any other change in how work gets done.
What you need to do now
Add AI and digital technology to your hazard register and risk assessment process, if it isn't already there.
Review any AI, automation, or monitoring tools currently in use, not just the ones you're planning to introduce.
Build consultation into your rollout process, before the technology goes live and on an ongoing basis.
Make sure workers get proper information, training, and instruction on how the technology will be used and how to raise concerns.
Revisit your codes of practice references. SafeWork Australia points to the existing model Codes of Practice on managing WHS risks and managing psychosocial hazards, both of which apply here.
How Anzen can help
This is exactly the kind of change that trips businesses up, not because leaders don't care, but because WHS systems haven't caught up with how fast the technology is moving. We help businesses translate guidance like this into practical, workable processes: reviewing what you've already introduced, building AI and digital technology into your risk management systems, and making sure your consultation processes actually meet the standard.
If you're not sure where your business sits with this, get in touch. We'll help you find out, and fix it, without the paperwork overload.
