Facebook and Instagram to train AI on your public posts in Europe. Can you opt out? Get information about Meta's new policy and how it affects your data privacy.
 |
| Meta Leverages Public Posts in Europe for the implications of using user's public data for AI development. |
Meta, the parent company of Facebook and Instagram, has sparked privacy concerns in Europe with its announcement that it will soon begin using public posts from European users to train its artificial intelligence (AI) systems.
While Meta assures users this practice complies with General Data Protection Regulation (GDPR) guidelines, the move has raised questions about transparency and user control.
The new policy will take effect on June 26, 2024, and will incorporate public posts, photos, captions, and messages sent to an AI assistant (excluding private messages) into Meta's AI training datasets. This data will be used to "develop and improve AI at Meta," according to a company statement.
Critics argue that using personal data, even publicly shared information, for AI training should require explicit user consent. Meta maintains that its practice falls under the GDPR's "legitimate interest" provision, but some European data protection agencies have expressed reservations.
The Irish Data Protection Commission (DPC) reportedly raised concerns with Meta regarding the policy and received clarification from the company. Meta has assured the DPC that only public posts from EU users and those over 18 will be used in the training process.
While users will have the option to opt out of their data being used for AI training, the specific method for doing so remains unclear. Meta has stated it will review opt-out requests, but the process is not yet available.
Additionally, even if a user opts out, their data could still be included if it appears in a friend's post or caption.
This move reflects Meta's growing focus on AI research and development. The company has invested heavily in this area, recently releasing its latest large language model (LLM) called Llama 3.
This model is designed to power AI assistants like Meta AI, which is not yet available in Europe. Meta has reportedly used public posts from Instagram and Facebook in the past to train AI assistants like Meta AI.
The news comes amidst a broader debate about the ethics and privacy implications of AI development. As AI becomes increasingly sophisticated and plays a larger role in our lives, ensuring transparency and user control over how our data is used will be crucial.
Meta's approach of using public posts for AI training, while GDPR-compliant according to the company, raises concerns and highlights the need for ongoing discussions about responsible AI development.