Every so often a piece of technology arrives that forces an industry to look in the mirror, and for Australian journalism that moment has turned up in the unlikely shape of a chatbot called Zack. Pitched as an artificial intelligence that can behave like a reporter, Zack recently sat down, so to speak, for a conversation with a writer at The Australian, and the resulting column has become a small but telling case study in how the profession is grappling with automation.
The premise is simple enough. Rather than using AI only to transcribe interviews, draft headlines or sift through documents, the people behind tools like Zack want the machine to attempt the whole act of reporting: gathering facts, asking questions, assembling a story and presenting it with the confident voice readers expect from a byline. It is the difference between an assistant that hands a journalist a sharpened pencil and a system that tries to be the journalist. That ambition is what makes the experiment interesting, and it is also what makes it unsettling.
From helpful tool to would-be reporter
For most newsrooms, AI is already a quiet fixture. It tags photos, suggests search terms, flags trends in large data sets and speeds up the grunt work that used to eat a reporter’s afternoon. Those uses are broadly accepted because a human still signs off on the final product. What Zack represents is a step across a line that many editors have been reluctant to cross, where the technology is no longer in the background but is auditioning for the job itself.
The appeal to publishers is obvious. Media companies have spent two decades watching advertising revenue drain away to global platforms, and the promise of a system that can produce serviceable copy around the clock, without a salary, superannuation or sick leave, is seductive to any chief executive staring at a shrinking budget. A bot does not get tired on election night, does not need a visa to cover an overseas story and can generate a hundred versions of the same market report before breakfast.
Yet the column’s encounter with Zack also exposed the limits of the pitch. An AI can mimic the cadence of reporting, but it does not knock on doors, cultivate sources over years, read a room or sense when an official answer does not add up. It cannot be held legally accountable in the way a masthead can, and it has no professional reputation to protect. When a machine confidently states something that turns out to be wrong, there is no reporter to ring back, no editor to issue a correction with their name attached, and no person who carries the consequences.
Two ways to read the experiment
Optimists in the industry argue that tools like Zack will free journalists to do the parts of the job that actually matter. On this view, if a machine can churn out routine summaries of council agendas, sports scores and company announcements, then human reporters can spend their time on investigations, analysis and the kind of on-the-ground work that no model can replicate. The technology becomes a force multiplier for a thinned-out workforce rather than its replacement.
The sceptics are not convinced, and their worry is less about quality than about incentives. Their fear is that cash-strapped owners will not use the savings to fund more investigative reporting but will simply cut costs, filling pages and feeds with plausible, automated filler while quietly letting experienced staff go. There is also the harder-to-measure risk to trust. Readers already struggle to tell a sponsored post from a news story, and a flood of machine-written copy that imitates journalism without its accountability could erode public confidence at exactly the moment the industry can least afford it.
Both camps tend to agree on one point: the technology is improving fast, and pretending it will go away is not a strategy. The question is whether the profession sets the terms of its use or has them set by whoever can deploy the cheapest system first.
Why this matters for Australia
Australia has a particular stake in how this plays out. The local media market is unusually concentrated, dominated by a handful of large players, and it has already shed thousands of journalism jobs over the past decade as regional papers closed and metro newsrooms shrank. A technology that can plausibly stand in for entry-level reporting lands in a sector that has little slack left to absorb another wave of cost-cutting. The cadet jobs that once taught young reporters their craft are precisely the roles an automated system is most likely to swallow, which raises an awkward question about where the next generation of senior journalists will come from if the training ground disappears.
There is a policy dimension too. The federal government has spent recent years trying to force global platforms to pay for the news they distribute, most visibly through the News Media Bargaining Code, on the logic that quality journalism is a public good worth protecting. If a meaningful share of that journalism is produced by machines, the argument for special treatment becomes harder to make, and the debate over who should pay for news gets tangled with the question of who, or what, is actually producing it. Regulators are also still working through how Australia’s defamation and consumer laws apply when an AI publishes something false, an area where the rules were plainly written with humans in mind.
For audiences, the practical upshot may be subtle rather than dramatic. The front-page investigation that embarrasses a minister will almost certainly stay human for the foreseeable future. It is the vast middle layer of everyday coverage, the market wraps, weather explainers and routine announcements, where a system like Zack could quietly become the default without most readers ever noticing the switch.
What’s next
Expect the experiments to multiply. More Australian publishers will trial AI reporters in low-stakes corners of their output, watching closely to see whether readers accept the results and whether advertisers care who wrote the words. The more consequential decisions will happen off the page, in the newsroom policies that spell out when a machine can and cannot carry a byline, how its work is labelled and who takes responsibility when it gets things wrong. The media union and press bodies are likely to push hard for disclosure rules so that readers always know whether a person or a program produced what they are reading.
Zack, in that sense, is less a threat than a prompt. A chatbot trying to be a journalist is a mirror held up to the profession, asking it to decide what is genuinely irreplaceable about the work and what was only ever busywork. The answer Australian newsrooms give over the next few years will shape not just their budgets but the trust they depend on to exist at all.
Sources: The Australian.
