Five ways to spot AI misinformation before you vote
Australian political parties are turning to AI "ensloppification" to win votes. We've seen One Nation post AI-generated, race-based Facebook clips attacking migration. Labor has shared a Halloween-themed reel using a cloned voice to claim Victorian Liberal leader Jess Wilson wants "Americanized" health care.
Five distinct methods exist to identify AI-generated misinformation before you cast your vote. In Australia, political parties are employing AI to craft misleading content for elections. One Nation posted AI-generated Facebook clips promoting racial division, while Labor utilized a cloned voice in a Halloween-themed video to accuse a Liberal leader of advocating Americanized healthcare.
This type of misinformation, known as misAInformation, aims to sway public opinion. Misinformation and disinformation both involve the spread of false content, but misAInformation is specifically produced using AI. Recent research indicates that AI-generated ads only recently emerged during Australia's 2025 federal election, coinciding with the rise of misleading claims in traditional media.
Digital platforms have yet to apply consistent rules for disclosing generative AI use, but they claim to have strengthened labeling and detection efforts. MisAInformation employs familiar tactics, such as twisting facts, exaggerating claims, and creating or altering images, video, and audio. Fact-checking has long been a crucial aspect of election integrity, but the sector has faced challenges since 2025.
Meta terminated its third-party fact-checking program in the U.S., citing trust issues, while other platforms have reduced their fact-checking efforts. The Victorian Electoral Commission's "Stop and Consider" campaign offers a four-question guide to help discern fact from fiction. To uncover AI-generated misinformation, consider these five strategies.
Firstly, investigate whether the content has been presented in a previous context using Google Reverse Image Search. Secondly, trace the origin of fabricated media, looking for original sources and verifying if any synthetic elements have been added. Thirdly, utilize digital provenance standards like those endorsed by the Coalition for Content Provenance and Authenticity (C2PA), which tag photos, videos, and audio with Content Credentials, indicating creation details and editing history.
If any tampering occurs, it is evident. Fourthly, seek invisible watermarks as a supplementary verification method. Lastly, employ AI detection tools, but remain cautious since they often produce inaccurate results. When in doubt, consult trusted professionals like AAP FactCheck, ABC News Verify, Crikey, or other Australian news outlets with verified teams.
Finally, for election-related content, consult the Victorian Electoral Commission's misinformation register, which debunks common election myths. While AI alters political campaigning, the ultimate decision at the ballot box remains yours. Make sure your vote is informed by facts.
Written by urgent.news from Phys.org's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.