One Nation Screens AI-Generated Video Full of Racist Tropes at Campaign Launch

One Nation has screened an AI-generated campaign video at its Victorian campaign launch — a clip that depicts a man who appears to African brandishing a machete, an apparently Middle Eastern man wearing an explosive vest, and a South Asian man in a turban discussing "mass migration."

The video, reportedly played at the party's launch over the weekend of October 3, 2026, has drawn immediate condemnation. Critics describe it as a new low for a party with a long history of controversy. The clip also includes an AI-generated Indigenous elder figure who speaks about people receiving "money … on the basis of need, not identity" — language that echoes the party's stated campaign positions against treaty processes and identity-based programs.

The video is vile and tired, as critics have put it. But the deeper problem it exposes is bigger than One Nation.

What does the video actually show?

According to reporting on the launch, the video is a stitched-together sequence of AI-generated characters, each built around a racial stereotype:

• A man who appears to be African, depicted brandishing a machete.
• An apparently Middle Eastern man, depicted wearing an explosive vest.
• A South Asian man in a turban, talking about "mass migration."
• An Indigenous elder figure, talking about people receiving "money … on the basis of need, not identity."

The clip aligns with policy positions One Nation promoted at the same Victorian launch, including its push to abolish First Nations treaty processes — positions the party laid out publicly at the event.

This is not subtle political messaging. It is a montage of fears, generated by a machine, assembled to provoke. And it was shown at an official campaign event, not posted anonymously by a fringe account.

Why does this moment matter for AI and politics?

The One Nation video lands in a regulatory vacuum. Generative AI tools can now produce photorealistic video of people who do not exist, saying things they never said, in minutes and at negligible cost. That capability has outpaced the rules governing political communication in Australia.

The result, as commentary on the clip has noted, is a "wild west of online politicking" — a space where unrestricted access to AI tools lets any actor, from a major party to an anonymous troll farm, manufacture racist or deceptive content and distribute it at scale during an election campaign.

The timing sharpens the stakes. The video surfaced in early October 2026, during the run-up to Victoria's state election — a period when voters are most exposed to campaign material and when fabricated content can spread fastest.

Australia's election and media watchdogs have limited powers over this kind of material. Existing frameworks were built for television ads, printed pamphlets and human-authored social media posts. They were not built for synthetic video that can be generated, tweaked and re-uploaded faster than any regulator can review it.

Is this an isolated incident or a pattern?

Commentators who reviewed the clip argue it is part of a broader pattern — what one critic labeled the "ensloppification" of Australian politics. The term captures a specific dynamic: AI tools produce cheap, high-volume, low-quality content, and political actors deploy it because it costs almost nothing and reaches audiences through algorithmic feeds that reward provocation.

Under that dynamic, the constraint on political communication is no longer production cost or craft. It is only the willingness to publish. A party that once needed a production budget, actors and a broadcast slot to put a racist video in front of voters now needs a prompt and a weekend.

The One Nation clip demonstrates both halves of the problem. It is technically crude — critics called it "tired" — and it is nonetheless circulating in an official campaign context, where its racism does the political work regardless of its production quality.

That is the core shift generative AI has delivered to democratic communication: the floor for credible-looking political media has collapsed, and the incentives favor whoever is least constrained by norms.

What are watchdogs asking for?

Reporting on the video concludes that Australia's watchdogs need more powers to do something about AI-generated political content. Under current rules, there is no clear, fast mechanism to pull a synthetic campaign video — even one built on openly racist tropes — from circulation during an election period.

The gaps identified in commentary on the clip include:

• No effective pre-emptive takedown authority for synthetic political content during campaigns.
• Rules written for traditional media that do not address AI-generated video of fake individuals.
• Enforcement timelines that lag the speed at which AI content spreads on platforms.

None of those gaps are hypothetical now. The One Nation video is a live, documented example of AI-generated racist content screened at a major party's official campaign launch — precisely the scenario regulators and researchers have warned about as generative video tools went mainstream.

Who is accountable when the speaker doesn't exist?

The video raises a question Australian law has not fully answered: when an AI-generated character delivers a racist message, the standard accountability chains — the candidate who said it, the spokesperson who stands behind it, the broadcaster who aired it — all weaken or break.

One Nation did screen the video at its own launch, which anchors responsibility with the party in this instance. But the same tools that produced the clip can produce content with no such anchor: anonymous accounts, unattributed ads, synthetic "vox pops" that no human ever performed.

That is why commentators frame this clip as evidence for a structural fix rather than a one-off scandal. Condemning One Nation addresses this video. It does not address the next ten videos, from any actor, in any campaign, at any level of Australian politics.

What happens next?

The video's emergence during the Victorian campaign puts pressure on two fronts. First, on One Nation, which must answer for screening openly racist, machine-generated stereotypes at an official event. Second, on Australian regulators and legislators, who now have a concrete, high-visibility case demonstrating that unrestricted AI access has arrived in federal and state political campaigning.

How watchdogs and platforms respond during this campaign will set the reference point for every Australian election that follows — because the tools that made this video are not going away, and the parties watching this episode unfold now know exactly what they can get away with publishing.