AI Deepfakes of Victorian Leaders Ignite Election Fight

Victorian Labor posted an AI-generated video on Saturday that shows Liberal leader Jess Wilson saying words she never said, opening a deepfake flashpoint in the state's election campaign weeks before voters go to the polls.

The video, published to Labor's social media accounts on Halloween night, depicts Wilson trick-or-treating and warning of "Americanised healthcare" under a future Liberal government. "If you had a heart attack, you would have to pay $150,000 for care," the fabricated Wilson says, in a voice engineered to sound like hers.

The depiction shows Wilson clutching a bill itemising the supposed costs of US-style healthcare: $150,000 for a heart attack, $6,000 for an ambulance and $40,000 to have a baby. Beside her stands One Nation's Victorian leader, Warren Pickering, wearing a Liberal party badge.

Wilson never made those statements. The claims trace back to a press conference last month, after she announced a $1.27 billion plan to pay private hospitals to conduct public elective surgeries to reduce the waiting list. Asked whether it was bad to "Americanise" the state's healthcare system, Wilson answered: "Not at all."

"This is happening in Queensland today. In fact, the Labor government has implemented it here in the past," she said at the time.

How did the premier defend the video?

Premier Ben Carroll defended the video's content on Wednesday on ABC Radio Melbourne, arguing the attack rested on Wilson's own record. "Jess Wilson herself has said there is nothing wrong with Americanising our health system. They're her words. She has said that, and every Victorian should be terrified by that," Carroll said.

Wilson rejected the framing outright. She called the video a "desperate smear campaign" and insisted "in no way, shape or form are we Americanising the healthcare system here in Victoria."

The Liberal party campaign described the video as "garbage" and urged voters to treat it that way.

Are the Liberals using AI too?

The Liberals have not abstained from the same technique. Among the party's social media posts ahead of this month's election are AI-generated videos showing Premier Ben Carroll asleep at his desk. The videos followed a front-page Herald Sun story on 24 October, in which unnamed sources accused the Labor leader of taking naps on the job.

A source inside Labor's campaign headquarters, not authorised to speak publicly, said the Liberals "threw the first brick" by creating that video.

The practice extends beyond the two major parties. The conservative lobby group Advance Australia released an AI-generated video last month that fabricates a Labor victory on election night, complete with former premiers Jacinta Allan and a North Face-clad Daniel Andrews. A video shown at One Nation's campaign launch conjured inflammatory tropes of people from various ethnic backgrounds.

Only the Advance video carries a disclaimer, stating that "this nightmare version of election night has been imagined by AI."

What do researchers say about the risk?

The wave of synthetic campaign material has led experts and the Greens to call for the practice to be banned altogether.

"Trust is already at all-time lows," said Clancy Moore, chief executive of Transparency International Australia. "So lies, deepfakes and AI-generated politicians in election ads only further erode Victorian's trust in politicians and our democracy."

Timothy Koskie, a media researcher at the University of Sydney, notes that misinformation in political messaging on social media "predates AI pretty significantly." What has changed is the ease of manufacturing convincing visuals. Synthetic video and images make it easier to play on "some subtle mischaracterisations that are easier to represent artificially."

"It's not that we're deceiving voters into believing something that's not true. It's that we're making more convincing an argument that actually doesn't hold up," Koskie said.

He is pessimistic about voluntary restraint. "It's very difficult to stop it from happening. The public won't reward people who refrain from using it."

Koskie argues those spreading misinformation should be held accountable for "their mischaracterisation of information," and that accountability must arrive before the damage spreads. "Trying to clean it up after is a losing proposition," he said.

He points to enforcement economics as a workable model, citing X's $650,000 fine for failing to comply with an Australian child safety notice. Even small financial penalties can change behaviour, he argues.

"These little nudges have some pretty drastic impacts over time. So we don't necessarily have to drag these people over coals or torture them in public if they make a misstatement, but that misstatement needs to have some sort of a follow-up," Koskie said.

What laws already exist elsewhere in Australia?

Moore is calling for Victoria and the federal government to introduce truth in political advertising laws, which currently exist in the ACT and South Australia.

South Australia's laws ban the use of AI-generated deepfakes to create electoral ads that depict acts a person did not perform, unless that person grants permission. Penalties reach $5,000 for individuals caught in breach.

The South Australian framework does allow a person to consent to their image appearing in a deepfake advertisement. But those advertisements, and any others that use AI, must be labelled as such.

Moore says that at a minimum, political parties in Victoria should apply an AI watermark to inform voters what they are watching. On that measure, the record so far is poor: of the videos circulating in this campaign, only Advance Australia's carries a disclosure.

What are the Greens proposing?

The Greens will on Thursday call on whoever forms government after this month's election to introduce South Australia-style truth in political advertising laws.

"Victorians deserve to know that what they're seeing and hearing from political parties is real," said the party's state leader, Ellen Sandell.

Sandell has also urged the parties to sign a memorandum of understanding setting out standards of behaviour during the campaign, including a commitment not to use "deceptive AI content."

Why does this election matter for AI regulation?

Victoria's campaign has become a live test of whether Australian political parties will police their own use of generative AI in the absence of binding state or federal rules. Both major parties have now deployed fabricated depictions of their opponents while condemning the other side's versions, and neither has committed to labelling.

The stakes extend beyond one state contest. With South Australia's $5,000 penalty regime and mandatory AI labelling already operating, and the ACT's truth-in-advertising laws in force, Victoria and Canberra face a concrete legislative template rather than an open question. Whether the next Victorian government adopts it will signal whether Australia's deepfake rules harden before the next federal campaign, or whether disclosure remains, as in this election, almost entirely voluntary.