A series of misleading and AI-generated attacks on Australian political parties on the Chinese social media platform RedNote (RED) has raised fresh concerns about misinformation, disinformation, and potential foreign interference in the lead-up to the federal election.
Among the most notable incidents is an AI-generated deepfake video of Opposition Leader Peter Dutton, falsely depicting him speaking Mandarin. The video, which uses deepfake technology to mimic Dutton’s voice and mannerisms, falsely claims that he intends to deport Chinese Australians.
Another post targeting the Albanese government falsely suggests that Australian police are arresting temporary migrants in shopping centres.
RED, known in China as Xiaohongshu (or “Little Red Book”), was originally designed for lifestyle content and e-commerce. However, with over 700,000 monthly users in Australia, analysts warn that its growing influence and algorithmic structure make it a powerful tool for political misinformation.
The RECapture project, an initiative led by several Victorian universities, has been tracking misleading content on RED since November. Their findings reveal that both major parties and independent candidates have been targeted in coordinated disinformation campaigns.
Peter Dutton has been a primary target, with misleading posts distorting his positions on nuclear energy, immigration, housing, and healthcare but Labor’s diplomatic efforts with India have also come under fire with videos posted suggesting it could lead to more competition for housing between Indian and Chinese migrants.
Official accounts of Australian politicians on RED also appear to have been ‘shadow banned’—meaning their content is not searchable on the platform. The ABC confirmed that Labor, Liberal, and Independent candidates all experienced visibility issues while unverified, misleading content about them spread unchecked.
The Australian Electoral Commission (AEC) has acknowledged the growing threat of AI-driven misinformation and foreign interference but remains constrained by limited resources and jurisdictional challenges.
An AEC spokesperson told the ABC: “If we receive information about unauthorised content on the channel, we will have a look at it and if it requires action, we will take it. This is the standard approach we take across all social media platforms,” an AEC spokesperson told the ABC.
This reactive stance contrasts with efforts on mainstream platforms like Facebook and Instagram. Unlike Meta, which has introduced fact-checking partnerships, AI disclosure policies, and election integrity measures across Facebook and Instagram, the AEC lacks the same real-time monitoring capabilities on platforms like RedNote (RED).
Analysts warn that this lack of oversight on RED could leave voters vulnerable to manipulated content ahead of the election.
