A Halloween AI video targeting two Australian politicians reopens the debate on deepfakes in campaigns

By Yahoo News Australia and Herald Sun · November 4, 2026.

What we know is limited, and that is worth stating. Yahoo News Australia reports in its headline that a premier is “under fire” over an AI deepfake video. Herald Sun refers to a “deepfake shame”: according to its headline, the Labor Party is facing criticism over a Halloween AI video targeting Jess Wilson and Warren Pickering. We have not been able to read the text of either piece: the downloaded Yahoo page showed only the cookie notice, and for Herald Sun we have only the headline. We do not know the content of the video, who made it or who shared it. Nor do we know whether there has been an official response. All of the above is attributed to those outlets' headlines, not facts we have verified.

Our take is that the story matters less for the specific video than for what it signals. Generative AI is no longer just a tool for anonymous scammers. It is entering ordinary political communication, even in a humorous wrapper like Halloween. Therein lies the risk. A deepfake labeled as a joke can circulate without the label, and someone watching it on a phone rarely distinguishes satire from impersonation.

More broadly, regulation of political deepfakes is moving more slowly than the technology. The useful approach is not to ban satire but to require clear labeling and accountability from whoever publishes. If a political party uses these tools against its rivals, it normalizes a practice that will eventually turn against everyone.

In the short term, there will be more episodes like this, and trust in video as evidence will erode. In the long term, content authentication and provenance tools, along with clear campaign rules, may contain the problem. Until the details are published, the prudent course is not to draw conclusions about who is to blame.