Meta has confirmed that it scrapes public photos and posts from Australian users to train its artificial intelligence models, providing no mechanism for individuals to opt out. While the company claimed initially that it had not processed data dating back to 2007, it later admitted that all public content from that period on Instagram and Facebook is included in its training datasets, provided the user has not manually set the posts to private. This practice applies to adult accounts, raising significant concerns about the extent of data harvesting without explicit consent. The disparity in user rights highlights a critical double standard in Meta’s global privacy strategies. Unlike users in the European Union, who are granted an opt-out option due to strict regional regulatory requirements and legal uncertainty, Australian citizens are forced into default consent. Meta’s leadership acknowledged that this differential treatment is a direct result of differing legal landscapes, effectively admitting that stronger privacy protections in Europe would have shielded local data from being used for AI development. This revelation is highly relevant to the open_data movement, as it underscores the urgent need for robust legal frameworks to protect personal information in the age of AI. The incident illustrates how the absence of stringent privacy laws allows tech giants to exploit public data for commercial gain, disproportionately affecting vulnerable groups, including minors whose images are captured on family accounts. It serves as a powerful argument for updating privacy legislation to ensure transparency and control, preventing the unchecked monetization of personal digital footprints.
Source: abc.net.auPublished on 2024-09-11