When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
Facebook has acknowledged that it trains its AI on the public data of every adult user in Australia with no opportunity to opt out.
At an inquiry, Meta global privacy director Melinda Claybaugh told senators it was “correct” that the company had been scraping data from Facebook since 2007 as well as from Instagram.
Private posts and those from users under 18 weren’t included. However, Claybaugh said that Facebook would have scraped photos of children on adult users’ accounts. It wasn’t clear if this included data from people who turned 18 over that period.
Meta said in June that it would rely on users’ data in the European Union and US to train generative AI if they didn’t opt out. However, the company halted its AI launches in Europe over a “lack of certainty” due to regulations, Claybaugh said. Meta has criticized the EU for allegedly hurting innovation by limiting how it can train data.
Greens Senator David Shoebridge, however, argued that EU laws like the GDPR and AI Act kept privacy “protected.” He added that people in Australia would have been shielded against Facebook data scraping if similar legislation had been in place.
Such laws could come soon. The Australian government is close to unveiling promised reforms to its Privacy Act following a 2020 review that determined the law was out of step with modern times. Leaders have already introduced voluntary AI guidelines and are asking for input on making them mandatory.
The inquiry and any upcoming legislation could impact not just Meta, but fellow generative AI heavyweights like OpenAI, Apple, and Google. As in the EU, they might have to either delay rollouts (as Apple has done with Apple Intelligence) or limit the amount of region-specific training data.
