Meta-AI: How to protect your data from access!

Transparenz: Redaktionell erstellt und geprüft.
Veröffentlicht am

Meta plans to use data from Facebook, Instagram and WhatsApp for AI training from May 27, 2025. Users can object.

Meta plant, ab 27. Mai 2025, Daten von Facebook, Instagram und WhatsApp für KI-Training zu nutzen. Nutzer können widersprechen.
Meta plans to use data from Facebook, Instagram and WhatsApp for AI training from May 27, 2025. Users can object.

Meta-AI: How to protect your data from access!

Meta plans to use publicly available content from Facebook, Instagram and WhatsApp to train its artificial intelligence (Meta-AI) from May 27, 2025. This causes many users to worry about their data integrity. Those affected have the opportunity to object to this data use until May 26, 2025, as 5min.at reports.

Objection to the use of data must be made via a form that requires registration with Facebook or Instagram. However, it is not necessary to give a reason for this. Users who submit their objection on time will receive a confirmation email. What is interesting is that the planned data for AI training not only affects future content, but also all previously publicly available information on the platforms. Given the long existence of Facebook and Instagram, the amount of potential data is significant.

Possible contradictions and uncertainties

Once a data set is used for AI training, it cannot be deleted or retrieved, further reinforcing the concerns of data protection stakeholders. The situation with WhatsApp is particularly critical. As the consumer advice center states, there is no possibility of objecting because the chat content is end-to-end encrypted. Only messages sent directly to the “Meta AI” chatbot could be used for training purposes. As long as no data is sent to the chatbot, Meta has no access to the chat content.

Users should be aware that Meta's privacy policy states that for linked accounts, once an objection has been expressed, it does not need to be repeated. However, with unlinked accounts, each action is required individually. Even if an objection has already been made, it is recommended that you raise it again to be on the safe side, especially since there are no reports of rejected objections to date.

If the objection is rejected, users have several options. This includes, among other things, deleting existing content, deleting your own accounts on Facebook and Instagram and lodging a complaint with the data protection authority. Alternatively, legal action against Meta may be considered. Users should also take action if, despite an objection, they come across personal data that appears in AI-generated content.

Active measures for users

If you do not have access to Facebook or Instagram, you can still object to the use of your data. In addition, Meta offers the ability to request data from third parties or revoke their use. In this context, it is important to note that filling out a form is necessary to access this data or request correction or deletion. The specific “prompts” that led to the generation of the content must be specified.

The development and use of artificial intelligence raises numerous questions about data protection. The Baden-Württemberg data protection website also offers comprehensive information on the relevant legal bases and how to view personal data. Users should make informed decisions and act in a timely manner to protect their data rights and prevent possible abuses.