Maxaa dhacay
The Guardian reports that Fix Victoria and Better Victoria are using AI-generated social-media advertisements to attack Labor policies and emphasize crime, corruption, debt, roads, schools and hospitals. Fix Victoria’s advertisements include dramatic scenes involving crime, ambulance delays, potholes and a flood representing Victoria’s projected debt.
The Guardian reports that Fix Victoria spent just under A$100,000 on Google and Meta advertisements in August, while Better Victoria spent about A$40,000, based on tracking by advertising executive Ed Coper’s agency, Populares. The report says the groups spent more on those platforms than any state election candidate or party during that month, and that Fix Victoria also bought television advertising during a Channel Seven AFL broadcast.
According to The Guardian, Fix Victoria was registered with the Australian Securities and Investments Commission in early August by Deborah Henderson, a former deputy executive director of the Institute of Public Affairs. Its communications director Andrew Hudgson and head of advocacy Gideon Rozner also previously worked at the conservative thinktank. The organisation says it is separate from the IPA and is “issues-driven, not politics-driven.”
The Guardian reports that Better Victoria was formed about a month after Fix Victoria. Its secretary is listed as Melbourne lawyer Ronald Holzer, who said he has only an administrative role. An unnamed spokesperson said the group had no relationship with any political party and was funded by member donations, but declined to identify its operators or donors.
The Guardian reports that Victorian electoral law does not require groups to register as third-party campaigners unless they specifically promote or oppose a party or candidate. The Centre for Public Integrity said this leaves a disclosure gap, while Transparency International Australia said the advertisements were clearly intended to influence voting. The groups say they have obtained legal advice and comply with relevant laws.
Faahfaahinta isha: theguardian.com ↗
Maxay muhiim u tahay
The report highlights how AI-generated political advertising can make emotionally powerful campaign messaging cheaper and faster while electoral rules may leave voters unable to identify its funders. That combination matters because voters may see persuasive synthetic imagery without knowing who commissioned it, what interests support it or whether the claims in the ads are accurate.
The Guardian’s reporting shows a practical accountability problem rather than establishing that the advertisements are illegal. Groups can frame messages around issues such as debt or public safety without explicitly naming a candidate or party, potentially avoiding disclosure requirements while still using campaign-style material.
AI is directly relevant because the reported advertisements rely on synthetic, high-impact scenes to communicate political claims. However, The Guardian does not independently verify the ads’ effectiveness, the factual accuracy of every claim, the identity of the donors or the extent to which AI changed the groups’ production costs. The article also does not establish that the ads violated platform rules or electoral law.
Farsamaynta Is-dhexgalka: Sida Dhabta Ay U Shaqeyso
U baadh tignoolajiyada hoose ee ka dambeeya horumarkan si isdhexgal leh.
Why can ethical evaluation not be reduced to one model score?
Maxaa la daawan doona xiga
The main unknowns are who ultimately finances the groups, how their advertisements are targeted, whether platforms label the material as AI-generated, and whether the campaign changes voter behavior. Watch for disclosures by the groups, scrutiny from electoral authorities and any proposals to broaden third-party campaigner rules.
Watch whether Fix Victoria or Better Victoria voluntarily identify their financial backers or political affiliations, and whether Victorian regulators assess whether their activities meet the legal definition of third-party campaigning.
Monitor whether Google, Meta or Channel Seven disclose additional information about the advertisements, including targeting, reach, labeling or content-review decisions. The report does not document general access to all ads or their targeting data.
The report’s spending figures are estimates based on public advertising data and should not be treated as independently audited totals. No evidence in the source demonstrates a measurable effect on voting intentions or election outcomes.