SAO PAULO (Reuters) – Brazil's National Data Protection Agency (ANPD) has decided to immediately suspend the validity of Meta Inc's new privacy policy on the use of personal data to train generative artificial intelligence systems in the country.
The ANPD's precautionary measures, published in Brazil's Official Gazette on Tuesday, suspend the processing of personal data across all Meta products, including data of people who are not users of the tech giant's platforms.
Authorities under Brazil's Ministry of Justice said non-compliance would be punished with fines of 50,000 reais ($8,836.58) a day.
In its ruling, the ANPD cited “an imminent risk of serious, irreparable or irreparable damage to the fundamental rights of affected rights holders.”
According to Brazilian authorities, Meta must revise its privacy policy to remove the section about processing personal data for generative AI training, and provide a public statement that it has suspended processing personal data for that purpose.
Meta said in a statement that he was “disappointed” with the ANPD's decision, adding that the measure represents a “setback for innovation” and will delay the benefits of AI for Brazilians.
“We are more transparent than many companies in our industry who use public content to train their models and products,” Meta said. “Our approach complies with Brazilian privacy laws and regulations.”
(1 dollar = 5.6583 reals)
(Reporting by Paula Arend Leyer; Editing by Anil DeSilva)
