- Filed
- September 9, 2026, 04:10 UTC
- Actor
- Sri Lanka-based commercial influencer / meme-factory network (Reset Tech attribution; operating globally under identical playbooks)
- Target
- Australian Facebook users and political discourse; secondary audiences in France, Italy, UK, US, Canada, Mexico
- Confidence
- high Multiple independent open sources, low disagreement about the facts
- Techniques
- Synthetic mediaAstroturfingCard stackingFlooding the zone
- Channels
- Facebook pages and Reels
- AI-generated deepfake video and imagery
- Meta creator monetisation
- engagement-bait polls and outrage content
Summary
Reset Tech research, reported by ABC News on 9 September 2026, documents a coordinated set of more than 200 Facebook pages that generate and post AI deepfake videos and images of dozens of Australian politicians. The network is linked to a Sri Lanka-based influencer business, operates identical branding across multiple countries, and monetises roughly one in ten pages via Meta’s creator payments. Content centres on inflammatory, anti-migrant or sexualised themes designed to maximise clicks rather than advance a coherent state ideology.
Analysis
Primary evidence is the Reset Tech investigation summarised and visualised by ABC News (Ange Lavoipierre, 9 September 2026). Researchers tracked more than 200 Facebook pages over eight months that systematically impersonate Australian political figures including Anthony Albanese, Angus Taylor, Pauline Hanson, Fatima Payman, David Pocock, Allegra Spender and Tony Burke. Short Reels depict the politicians delivering fabricated inflammatory statements—anti-migrant, anti-Muslim, or sexualised content—often ending with bait polls. Pages adopt mass-appeal names such as I Love Australia or Life in Australia, seed audiences with non-political AI images of animals or scenery, then pivot to political deepfakes. Approximately 10 percent of the pages are monetised through Meta’s in-house creator payments, generating direct revenue for the operators.
The network is attributed to a Sri Lanka-based influencer business that runs parallel operations under the same branding templates in at least six other countries, with a global footprint of 1,864 coordinated pages and 5.9 million combined followers. Only one in three deepfake posts received AI labels and one in 33 was removed, despite clear violations of Meta’s policies on inauthentic behaviour, hate speech and synthetic media.
The core mechanism is commercial synthetic-media farming. Deepfakes transfer the visual authority of real politicians onto fabricated speech (transfer via synthetic media), while the coordinated page network and identical posting patterns constitute astroturfing of organic Australian political conversation. Selective fabrication of statements that maximise outrage constitutes card-stacking of the political record; the industrial volume across hundreds of pages floods the zone of available political content on Facebook. Because the decisive driver is revenue from engagement rather than a coherent ideological or state directive, the operation sits at the commercial end of the influence spectrum yet still distorts the information environment that citizens use to evaluate public figures.
This qualifies as an influence operation because origin is concealed, authenticity is artificially constructed, and the output systematically injects false political speech into domestic discourse at scale. Identical industrial deepfake and monetisation patterns appear in other commercial meme factories targeting elections or polarised topics; the technique set is therefore scored independently of any particular national actor.
Literacy counter: treat short-form video or image posts that show familiar politicians delivering highly polarising or out-of-character statements as provisional until verified against official channels or primary footage. Check for AI labels, reverse-search frames, and note whether the page name, creation date and posting cadence match known commercial networks. Prefer primary sources, parliamentary records and established local journalism over engagement-optimised Reels. Demand that platforms enforce existing synthetic-media and inauthentic-behaviour policies at the scale of the problem rather than relying on post-hoc researcher discovery.
Evidence
- ABC News: Foreign 'AI slopaganda' network targets Australian politicians with deepfakes (9 Sep 2026)
- ABC News audio companion report (9 Sep 2026)
- Prior ABC Verify reporting on foreign AI Pauline Hanson accounts (March 2026)
- Guardian Australia on foreign meme factories monetising pro-One Nation groups (June 2026)