Think about the last near miss you experienced at work. The ladder that wobbled but didn’t fall, the forklift that stopped just short or the comment that felt off but you let go. Did you report it?
If you are like most Australian workers, the answer is probably no. Research puts the share of safety incidents that go unreported in this country at almost 1 in 3. This isn’t limited to the frontline either, with around 1 in 4 leaders and managers underreporting too.
It is tempting to blame the clunky paperwork or a busy workload, but this can’t be the whole story if leaders are staying quiet at nearly the same rate as everyone else. It’s not about knowing the safe thing to do either, ask almost anyone and they will tell you they should report it. Something else is going on, more to do with how we are wired to weigh cost and benefit in the moment.
Daniel Kahneman won the Nobel Prize in Economics for his research into how we make decisions, later distilled into two ways we think. System 1 is fast, automatic and runs on instincts, with wiring far older than we are. System 2 is slow, effortful and deliberate, a more recent addition to our cognitive toolkit. Reporting an incident is a System 2 task: it takes time, it draws attention and it might invite scrutiny. Staying quiet is the System 1 default: automatic and running on the kind of snap judgement that once kept us alive when hesitation was the bigger risk.
Put System 1 and System 2 side by side and underreporting stops looking like a training or systems problem and starts looking like exactly what you’d predict from a rational person weighing an immediate personal cost against a delayed, diffuse benefit. We evolved to be wired for silence.
If you add that only a third of managers say every incident reported leads to corrective actions, the picture gets worse. Even the people who fight their own instincts to speak up are often learning that over time, it wasn’t worthwhile. This is a safety experience gap and it’s the kind that compounds.
So what actually closes the gap? Where does new technology, and AI in particular, genuinely lower the barrier to speaking up? Just as importantly, where does it fall short, no matter how well it’s built?
Faster reporting with AI
Our 2026 Australian Workplace Safety market research report found that 49% of managers worry that incidents or near misses aren’t being reported at all. Some of that comes down to effort, for instance, a near miss late on a Friday afternoon is competing with everything else on the to-do list, and the report usually loses that battle.
With Rapid Incident Reporting, a worker can scan the QR code on the wall or open the Rapid Worker app on their phone to report an incident anywhere, at any time, even when there’s no signal.
With AI-powered reporting, they can take a photo of the incident or risk (or describe it out loud, in the language they’re most comfortable with) and AI writes up the draft report for review. This means the worker stays in control of what the report says, and your safety team gets something they can work with that they otherwise might not have.
Anonymous reporting for psychosocial hazards
Bullying, harassment, unreasonable workloads, a customer that crosses the line: these are all examples of psychosocial hazards and they’re some of the hardest things to report. Raising one usually means naming someone, and sometimes that someone is a superior. In our research, 24% of workers said they’d personally seen a safety incident go unreported. This figure covers every kind of incident, from the ones that are easy to talk about through to the ones that can draw attention and scrutiny.
WHS duties in Australia cover psychosocial health, as well as physical safety, so organisations need to hear about these risks. Rapid Incident Reporting gives workers the option to report anonymously, so someone who would never otherwise put their name to a complaint about a superior can still get it on the record.
Anonymity does come with a trade-off, though. If you don’t know who raised a concern, you can’t follow it up with them or ask if they’re okay. That is why it’s best offered as an option, alongside named reporting, for the times where someone wouldn’t otherwise speak up.
Following up on reports
When someone reports a problem and never hears about it again, they’re less likely to make the effort next time around.
In Rapid Incident Reporting, the investigation, root cause and corrective actions all sit against the original report, with each action having someone responsible for it, a due date and managers notified about open actions. Closing the loop with the person who reported it is still up to you, and often that touchpoint to let them know the report has been received and action has been taken, is what gets the next report in, and builds a culture of safety and accountability.
Where people make the difference
Any reporting tool depends on trust. If someone’s seen a colleague blamed for speaking up, an easier-to-use form or an anonymous reporting option likely won’t change their mind. That kind of trust is built by leaders acting on reports, and being seen to do so.
AI doesn’t take judgement out of the process either. The worker confirms their draft before it’s submitted and someone still has to look at the evidence and decide what should change.
Closing the gap
If near misses aren’t reaching your safety team, it’s worth working out where the friction is. Sometimes it’s the time a report takes, sometimes it’s the worry of putting your name to something sensitive, or sometimes it’s the sense that nothing will come of it.
Rapid Incident Reporting is designed to make each of those easier. A photo or a few spoken words become an AI-drafted report that keeps the worker in control, with optional anonymous reporting giving people a safer way to raise sensitive issues. Once a report is in, every corrective action gets an owner and a due date, so it’s easier to follow through and tell people what changed because they spoke up.
