Meta suspends training of its AI in Europe

Meta has decided to pause training of its artificial intelligence (AI) models in Europe, models that are already available in other parts of the world.

• The company took the decision at the request of its main regulator, the Irish Data Protection Commission (DPC), which welcomed the stop.

Context. Noyb (European Digital Rights Centre) asked data protection agencies in 11 countries, including Ireland and Spain, to monitor the multinational corporation.

Problem. Meta plans to change its privacy policy starting June 26th. According to Noib, this will allow him to “use the personal content that users have been posting for years.”

• The organization believes that the company headed by Mark Zuckerberg could use this personal data of users. without asking their consent under the guise of “legitimate interest.”

• “Meta could bring its AI technology to Europe if it bothered to ask for such consent, but it appears to be doing its best not to,” they decry in a statement.

What does Meta say? To develop any AI, you must first train it, and Meta does this by only using content and data “that is publicly available” and published on Facebook and Instagram.

70% of communications students believe that the use of AI will make it difficult to distinguish between news and hoaxes.

• The company says it “complies with European legislation and regulations (on privacy and data protection)” and that they are even “more transparent” than their competitors (such as OpenAI).

• But he needs input from European users to adapt “to their language and culture” and optimize his experience using these tools, he says.

Sources

Meta statement on the decision to delay training of its AI models in Europe

Irish DCP press release on Meta’s decision

Information from Meta about the new privacy policy

Noib’s complaint about metapractices

Reuters