ANPD will call on X to demonstrate the use of user data to train AI

The National Data Protection Authority (ANPD) stressed in the article that it must summon users of social networks to train their artificial intelligence tools.

The ANPD launched an investigation in July this year to evaluate the use of personal data by X Corp (formerly Twitter), in training artificial intelligence (AI) models. The focus of the inspection is to ensure compliance with the General Data Protection Law (LGPD), which regulates the processing of personal data in Brazil.

The recent change to the platform’s terms of use, which was announced to come into effect on November 15, 2024, has raised concerns about the mandatory use of user data to operate artificial intelligence systems, without explicit consent. Following questions posed by our column, the ANPD has taken an official position on the situation:

“The ANPD launched, in July this year, an investigation to assess the LGPD’s compliance with the use of personal data for X Corp’s generative AI training. The investigation, taking into account the ANPD’s terms of reference, is focused on the announced changes to the privacy policy. To date, the company has responded to inspection requests For information.

Regarding the recent announcement of changes to the terms of use, the inspection will call on the company to provide clarifications, especially regarding the effects of the change on the protection of personal data of users of the platform.

If there is a risk of serious harm to the data protection or privacy of data subjects, other measures may be taken.”

Sanctions and potential implications in Brazil

Applying the LGPD requires that any processing of data for AI purposes, such as using user posts to feed learning systems, must be done transparently and with the clear and informed consent of the data subjects. Failure to comply with these requirements could result in heavy fines, as well as jeopardize X Corp’s operations in the country.

There is precedent that any collection and use of data for AI training must be explicit and have users’ consent, something that is not yet clear in the new terms.say digital law experts interviewed by the column. Sources at ANPD also spoke privately to the column that X could actually be subject to legal penalties and even fines, if the new practices are found to breach the LGPD and expose users’ data to unnecessary risk.

In 2023, ANPD had already notified Meta of similar practices, resulting in the temporary suspension of the company’s services until a compliance plan was adopted.

The importance of supervision

The current investigation strengthens the ANPD’s role as an active regulatory body in Brazil, which closely monitors the practices of big tech platforms and ensures that users’ privacy is maintained. The ANPD’s tough stance serves as a warning to other companies operating in Brazil, making clear that the country does not tolerate the misuse of personal information, especially in the context of new technologies such as artificial intelligence.

Speculation now revolves around X Corp’s response to the ANPD callout. As of the close of this report, Elon Musk’s platform has not officially commented on the issue.