Monash University, Automated Society Working Group
Grant round: 2019
Amount: $49,579
Facebook’s data-driven micro-targeting “dark ads” lack accountability and could enable the spread of disinformation, discrimination and harmful stereotyping.
Researchers from Monash’s Automated Society Working Group have developed a tool to attempt to make public what ads are being served to people online and how advertising is distributed across demographic groups.
Dark ads are visible only to those to whom they are delivered, meaning the content is not available for public inspection, the pattern of distribution is unclear and they tend to be short-lived and undergo constant transformation - making them difficult to track. Monash’s tool builds on a study conducted by US news outlet ProPublica which was designed to track political advertising on Facebook, and is believed to be the first to study all ads appearing in people’s Facebook newsfeeds. It allowed users to automatically donate their adverts into a database, which could then be collated and filtered by demographic categories.
Download the project report below.
Download: Monash_Dark Ads Report.pdf4.51 MB
Download: Unregulated and Segmented Dark Ads on Social Media7.71 MB