Today, we're releasing a new paper sharing key recommendations on AI that urge platforms to prepare for a new era of content moderation. #AI #SocialMedia #OnlineGovernance #Meta
— Oversight Board (@OversightBoard) September 17, 2024
Meta has admitted to scraping public data from all adult Australian Facebook and Instagram users to train its artificial intelligence models without providing an opt-out option. This revelation came during a senate inquiry where Meta’s Global Privacy Director, Melinda Claybaugh, testified about the company’s data collection practices. Ms.
Privacy concerns or progress? 🔒📈
Meta gets go-ahead to train AI using UK FB & IG posts 🤖🇬🇧📱 https://t.co/mXPmRwMQ2x
— One Green Bean (@onegreenbean) September 18, 2024
Claybaugh confirmed that Australian data from public posts, photos, and other information is collected to build Meta’s generative AI tools. She acknowledged that all public posts since 2007 could be used unless users had explicitly set them to private. Labor Senator Tony Sheldon questioned whether posts dating back to 2007 were used, and Greens Senator David Shoebridge emphasized that any public posts not set to private would be scraped by Meta.
While Ms.
Meta has announced that it's going to start training its AI tools on public Facebook and Instagram posts, comments and photos shared by members of the public in the UK.
What does that mean for users and their privacy? 🤔 https://t.co/nBLc7Qp83e
— Privacy International (@privacyint) September 19, 2024
Claybaugh confirmed that data from accounts of people under 18 is not scraped, she admitted that public photos of children posted by their parents are collected.
meta’s AI training with public data
#Meta said on Friday that it will push ahead with controversial plans to use millions of #UK Facebook and Instagram posts to train its artificial intelligence (#AI) technology. https://t.co/ugUKkrC2bm
— Solihull Updates (@SolihullUpdates) September 16, 2024
However, she could not clarify whether the company scrapes data from users who were minors when they created their accounts but are now adults. In June, Meta informed users in the European Union and the United States about their data being used to train generative AI products like Meta AI, allowing them to opt out. This option was provided due to legal uncertainties surrounding strict privacy laws in those regions.
However, Australians were not given this choice due to the lack of similar privacy regulations in the country. Senator Shoebridge criticized the government’s inaction, stating, “The government’s failure to act on privacy means companies like Meta are continuing to monetize and exploit pictures and videos of children on Facebook.” He emphasized the importance of updating Australian privacy laws to better protect users. The government is expected to announce long-awaited reforms to the Privacy Act following a 2020 review that deemed the current laws outdated.
Attorney-General Mark Dreyfus had indicated the legislation would be announced in August. As debates around privacy laws and AI usage continue, the disparity in how user data is treated across different regions underscores the need for stringent privacy protections. Facebook’s practice of scraping Australian user data without consent highlights significant privacy concerns, urging lawmakers to take swift action to safeguard the privacy rights of their citizens.
