SYDNEY – In February 2026, 13-year-old boy Austin Appelbee made global headlines after swimming for four hours to rescue his family who had been swept out to sea in Western Australia.
Days later, a story began to circulate online showing the country’s newly hailed hero alongside Australian anti-migrant firebrand Pauline Hanson, who presented him with a “life-changing gift” – a £50,000 (S$86,100) cheque to support his schooling.
But the story, which was widely shared on Facebook and TikTok and can still be found on both platforms, was fake.
Experts in Australia say countless other similar examples have emerged with growing regularity, as the nation confronts a wave of AI-generated content disguised as news.
The content includes mass-produced AI slop as well as “pink slime”, where ideologically or politically motivated content is presented as local news.
Brigid O’Connell, a PhD candidate at Deakin University who is researching local news and misinformation, told The Straits Times that AI-generated content which was promoted by politicians and political parties posed a “serious democratic risk”.
She said she did not know who created the article about Hanson and Austin and that “there is no evidence it has come from (Hanson’s party) One Nation”.
“People tend to trust ‘local news’ more than overt political communication, so when the real source is hidden, voters can’t properly judge the motives, funding, or reliability of what they’re reading,” she added.
Analysts say the AI-generated content in Australia is primarily created by both local and foreign actors seeking to influence Australian politics.
In addition, some creators seek to make money through platforms such as YouTube that pay creators of popular content or by diverting users to websites that feature advertisements.
In May 2026, the first-known instance of a content producer publishing entire news sites in Australia made up of AI-generated pink slime journalism emerged – a phenomenon that has been prevalent in the US.
Sites purporting to be authentic local Australian news websites with names such as The Mandurah Reader and The Bunbury Guardian targeted areas in Western Australia such as the small cities of Mandurah and Bunbury. Much of the content was lifted from ABC News and then regenerated by artificial intelligence, with fake names and biographies of reporters.
The sites were taken down after inquiries by ABC News. The person behind the sites told the broadcaster they were an “experiment”.
T.J. Thomson, an associate professor of visual communication from RMIT University, told ST that Australia, like the US, had a growing number of regional areas where local news services had shut or been downsized, leaving them vulnerable to being targeted by pink slime.
He said AI-generated content that republished legitimate news was usually designed to make money, but “if producers just create fake news without any basis in reality, it is more about trying to mislead and influence people politically”.
“AI-generated content is becoming more and more prevalent in Australia as the technology is becoming easier to use,” he said.
“Often it can be international actors who are trying to capitalise on the Australian market and make money here or influence politics.”
Thomson said it was hard to know where much of the content was generated but “a lot of it is automated with bots from overseas countries such as Russia”.
He said much of the content is aimed at “getting people to be cynical and mistrustful and questioning everything” and that promotion of anti-immigration sentiments was particularly popular.
Such content, he added, likely favoured the populist, anti-establishment messaging of Hanson’s One Nation party, which has been surging in the polls.
“I think it does benefit certain parties more than others – the parties that say they want to radically change what we’re doing,” he said.
The federal government has recommended that AI developers label and watermark content to ensure that people are aware that it has been made by AI.
Releasing a National AI Plan in December 2025 that attempted to address AI-related threats, Minister for Industry and Innovation Tim Ayres said in a statement: “By being transparent about when and how (AI) is used, we can ensure the community benefits from innovation without sacrificing trust.”
But experts say further measures are needed, such as European Union proposals requiring those publishing AI-generated content on matters of public interest to disclose that the text is AI-generated.
Thomson went a step further, saying Australian election authorities should consider introducing a ban on AI-generated content during election campaigns, which typically last about five to seven weeks.
But, for now, the best method of prevention may be for users themselves to proactively take steps to ensure that online content is authentic and credible.
O’Connell recommended looking for mastheads with contact details and editorial policies, reviewing the “About” page of websites, searching the reporter’s name and checking the URL to ensure it does not mimic the site of well-known outlets.
“Transparency in news is vital because a democracy only works when people can see who is speaking, what their interests are and how information is being produced,” she said.
