Meta is no longer allowed to utilize personal data from Brazil to train its artificial intelligence models, according to the country's data protection regulator (Autoridade Nacional de Proteção de Dados or ANPD). The ban was issued due to concerns about potential harm and inconvenience to people.
What Prompted the Ban?
According to The Verge, the decision resulted from Meta's May privacy policy modification, in which the company asked permission to train AI using publicly available Brazil-based data from Facebook, Messenger, and Instagram, including posts, photos, and captions.
It also comes in response to a study released by Human Rights Watch last month. The investigation revealed that LAION-5B, a major dataset for AI model training, has identifiable images of children in Brazil. This, said HRW, puts these youngsters in danger of being targeted by deepfakes and other forms of exploitation.
READ NEXT : Meta Faces $13 Billion Fine For Violating Digital Laws With its 'Pay or Consent' Data Privacy Model
Meta vs. Brazil
In a report by The Associated Press, ANPD warned that Meta's May policy might soon undermine users' basic rights in a way that would be difficult, if not impossible, to fix.
Brazil is a major market for Meta, with 102 million Facebook accounts alone, according to ANPD. Reportedly, Meta will be punished with a daily punishment of 50,000 reais (about $8,808) if it does not comply with the order within five business days.
Meta informed The Associated Press that its revised policy aligns with Brazil's privacy rules and regulations. The company also noted that the decision hinders competition in AI research and development, slows innovation, and puts off the day when Brazilians will reap the advantages of AI.
Meta pointed out that customers have the option not to have their data used to train AI. However, ANPD said there are excessive and unjustified obstacles making this process difficult.
Join the Conversation