The organisation that runs Australia’s national broadband network has delivered a blunt message from Hobart: the same artificial intelligence tools reshaping legitimate business are also handing criminals a way to operate faster, cheaper and at far greater scale than before.
Speaking in the Tasmanian capital, the chief of nbn warned that AI is lowering the barrier to entry for online offenders, allowing them to automate the grunt work of fraud and cyber intrusion that once required time, skill and manpower. According to Pulse Tasmania, the warning framed AI less as a single new threat and more as a force multiplier that makes existing crimes, from phishing to identity theft, dramatically more efficient.
Why the warning lands now
The concern is not abstract. Australians lose billions of dollars to scams and cybercrime each year, and the country’s telecommunications and network operators sit on the front line of that fight. nbn carries a huge share of the nation’s internet traffic, which gives its leadership an unusually clear view of how the volume and sophistication of malicious activity is changing.
What has shifted is the economics of crime. In the past, a convincing scam email had to be written, often by someone whose first language was not English, which is why clumsy grammar became a running joke and, usefully, a warning sign. Generative AI erases that tell. It can draft flawless messages in seconds, tailor them to a specific victim using scraped personal details, translate them into any language and produce thousands of variations to slip past spam filters. The result is that a lone offender can now run a campaign that would once have needed a small team.
The same logic applies further up the chain. AI can help write and refine malicious code, probe networks for weaknesses, and even power voice-cloning tools that let criminals impersonate a relative or a company executive over the phone. For defenders, the challenge is that the volume of attacks rises at the same time as their quality, which stretches both human analysts and the automated systems meant to catch them.
Two ways to read the threat
There are competing schools of thought about how alarmed Australians should be. One view, reflected in the nbn chief’s Hobart remarks, is that the scaling problem is genuinely dangerous and getting worse, and that governments, telcos and businesses need to move quickly to keep pace. On this reading, the asymmetry favours the attacker: a criminal only has to succeed once, while defenders have to succeed every time, and AI tips that balance further towards the offender.
A more measured counterview, common among security researchers, is that AI is currently amplifying old techniques rather than inventing entirely new ones. The core defences still work, including multi-factor authentication, staff training, patching and healthy scepticism about unexpected messages. In that framing, the same AI arms race is also arming the defenders, who are using machine learning to detect fraud patterns, flag anomalous logins and shut down malicious infrastructure at a speed no human team could match. The question is whether defensive adoption keeps pace with offensive adoption, and that race is far from settled.
Both positions can be true at once. AI may not have handed criminals a magic weapon, but by cutting the cost and effort of running scams to almost nothing, it lets far more people commit far more crime. Scale, rather than novelty, is the real story, and it is scale that overwhelms victims and investigators alike.
What it means for Australia
For Australia, the stakes are sharpened by geography and demography. A relatively wealthy, highly connected population makes an attractive target, and the country has already endured a string of high-profile breaches at major companies that exposed the personal data of millions. Every one of those leaked records becomes raw material for the kind of personalised, AI-assisted fraud the nbn chief is describing, because the more a scammer knows about a victim, the more convincing the approach.
Regional and remote communities, including much of Tasmania, carry particular risk. Older Australians and those with less exposure to digital security advice are frequently the ones targeted by phone and message-based scams, and they are also the households most reliant on the national network for banking, health and government services. That makes the choice of Hobart for this warning more than a matter of convenience. It underlines that the threat is national, not confined to the big east-coast capitals.
The federal government has been building out its response through its cyber security strategy, the work of the Australian Signals Directorate and the National Anti-Scam Centre, alongside new obligations on banks, telcos and digital platforms to disrupt scams before they reach consumers. Telecommunications carriers such as nbn are increasingly being cast not just as infrastructure providers but as active participants in blocking malicious traffic, and remarks like these signal that the sector expects its role in the fight to grow.
What happens next
The immediate task is a familiar one made more urgent. Businesses will be pushed to treat AI-enabled fraud as a board-level risk rather than an IT footnote, and to assume that the polished, personalised message is now the norm rather than the exception. For individuals, the old advice holds but needs reinforcing: verify unexpected requests through a separate channel, be wary of urgency, and remember that a perfectly written email is no longer a sign of legitimacy.
Expect the policy conversation to intensify as well. Australia is already weighing how hard to regulate AI, and warnings from network operators add weight to the argument that guardrails should account for the technology’s criminal uses, not just its commercial ones. Whether that translates into new obligations on AI developers, tougher scam-disruption rules or simply more investment in defensive tools will be one of the defining debates of the coming year.
For now, the message from Hobart is a straightforward one. The tools that promise to make honest work more productive are doing exactly the same thing for dishonest work, and the country’s defences will have to scale just as fast to keep up.
Sources: Pulse Tasmania
