Qantas is now in formal talks with the Australian Services Union over how AI is being used in workforce decisions. The union did not request a meeting. They demanded urgent talks. That distinction matters more than most HR leaders are giving it credit for.
This story, reported by HCA Magazine, is not about artificial intelligence. It is about what happens when workers feel that AI is being used to manage them rather than support them, and they decide they have had enough of waiting to find out.
What Actually Happened at Qantas
The Australian Services Union moved to initiate formal talks with Qantas after concerns emerged about how AI tools were being applied in workforce decisions. The union’s language was not collaborative. It was urgent. That tells you something important about the temperature inside that organisation before this became public.
When a union escalates to formal talks rather than raising concerns through normal channels, it usually means the normal channels were not working. Workers had been watching something happen, feeling something shift in how they were being managed, and not getting satisfactory answers. The formal demand was not the beginning of the problem. It was the point at which the problem became visible from the outside.
For the full story, it is worth reading the HCA Magazine piece directly. Qantas union clash is a warning shot for every HR leader deploying AI.
Why This Keeps Happening
This is not the first time an organisation has found itself in a difficult position because AI was deployed in ways that affected people without those people being brought into the conversation first. It will not be the last.
The pattern is consistent. A business identifies an efficiency opportunity. A tool is selected and implemented. Processes change. Then, sometime later, the people inside those processes start to notice that something is different about how decisions are being made, how they are being measured, or how their roles are being evaluated. They ask questions. They do not get clear answers. Trust erodes. And then, depending on whether the workforce is unionised or not, the response takes different forms. Sometimes it is a formal escalation like this one. Sometimes it is quiet disengagement. Sometimes it is turnover that takes months to connect back to the original cause.
The cost is real in all three cases. The repair bill is always higher than the conversation that was avoided.
The Difference Between AI That Supports and AI That Manages
There is a meaningful distinction that most organisations are not drawing clearly enough, and workers feel it even when they cannot articulate it.
AI that supports people makes their work easier, faster, or less repetitive. It gives them better information to make decisions with. It handles the low value tasks so they can focus on the work that actually requires human judgement. People generally accept this kind of AI once they understand what it does, because it is clearly on their side.
AI that manages people is different. It monitors performance metrics, flags behaviour patterns, influences rostering or staffing decisions, or feeds into how people are evaluated or rewarded. This kind of AI can be used responsibly, but it requires a completely different level of transparency and workforce trust to land well. When it is introduced without that trust, and without proper explanation of what the system is doing and why, people fill the information gap with their own conclusions. Those conclusions are rarely charitable, and they are rarely wrong to be cautious.
The Qantas situation almost certainly sits somewhere in this second category. The specific application matters less than the principle it illustrates. If your team is uncertain about whether AI is watching them, measuring them, or influencing decisions about them, and you have not addressed that uncertainty directly, you are already behind.
What the Businesses Getting This Right Are Doing Differently
The organisations that are deploying AI without ending up in formal union negotiations or dealing with trust breakdowns are not necessarily using better technology. In many cases they are using the same tools. The difference is in the order of operations.
They are talking to their people before the tools are deployed, not after. They are explaining what the AI does, what data it uses, and what decisions it does and does not influence. They are making commitments about how that will be governed and then honouring those commitments. They are creating channels for workers to raise concerns about how AI is affecting their roles, and those channels actually work.
This is not complicated in principle. It is just slower than most organisations want to move. The pressure to implement quickly is real. The competitive pressure is real. But the cost of a trust breakdown at scale is also real, and it lands on a timeline that is much harder to predict and manage than a phased rollout would have been.
The businesses getting this right are not the ones moving fastest. They are the ones talking to their people before the union does.
What Australian Business Leaders Should Take From This
If you are an Australian business owner or HR leader who is currently deploying AI, evaluating AI, or planning to deploy AI in any function that touches how people work or how they are measured, the Qantas situation is a direct signal worth paying attention to.
It does not matter whether your workforce is unionised. The dynamic is the same. When people feel that AI is being used against them rather than for them, and they have not been given honest, clear information about what is actually happening, the relationship between employer and employee changes in ways that take a long time to fix.
The question worth sitting with is not whether your AI deployment is technically compliant or legally defensible. Those things matter, but they are the floor. The question is whether your people know what the AI does, know what it does not do, have had a genuine opportunity to ask questions and get straight answers, and believe that their interests were considered in the decision to use it.
If the honest answer to any part of that question is no, that is worth addressing now. The window to do it on your terms, rather than in response to a formal demand, is still open. But it does not stay open indefinitely.
The technology is the easy part. The people are the hard part. Qantas is a reminder of what the hard part looks like when it is left too late.
Get in touch at www.xsiv.au/#form