JAKARTA After launching Meta AI in the European Union (EU) in March, Meta has now revealed plans for its Artificial Intelligence (AI) model data training. Meta AI will be trained with user data.

All public content shared on Instagram, Facebook, and other social media will be used as an AI training tool. This content includes posts and comments made by Meta's social media users in the EU.

"The interaction of people with Meta AI, such as questions and requests, will also be used to train and improve our model," the company said, quoted on Tuesday, April 15. "Producing AI on Meta (conducted) to better understand and reflect their culture, language and history."

This training will be carried out in the near future after Meta sends notifications to its application users or to user email. In the notification, Meta will explain the various types of data they will use.

Meta also provides options for using the data. If there are people in the EU who object to the data being used for AI training, they can refuse the special forms provided by the company. This form can be filled in at any time without time limit.

"We have made this rejection form easy to find, read, and use, and we will respect all the rejection forms we have received, as well as those that have just been sent," explains Meta.

The company led by Mark Zuckerberg also confirmed that they would not train Meta AI using private messages on various platforms. Thus, user privacy and security will be maintained.

In addition, Meta will not use public data from children and teenagers' accounts. They will only train their AI using accounts over 18 years of age that are not locked or do not use private mode.


The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)