The Rise of Meta AI and User Concerns

Meta AI is rapidly integrating into the daily digital lives of WhatsApp, Instagram, and messenger users.This AI assistant, identifiable by its distinctive blue and purple circle icon, is accessible through the search bar and within existing conversations. However, its integration raises significant questions about data privacy and user control.

meta’s Assurances vs. Skepticism

Meta claims that its AI can only access messages explicitly shared with the service and that these interactions will not be used to further train the AI. The company emphasizes its commitment to user confidentiality, with the AI assistant even assuring users in direct responses that it respects their privacy.

However, these assurances are met with considerable skepticism. Hélène Lebon, a lawyer specializing in personal data protection, voices concerns about Meta’s track record. She points to past instances, such as Facebook’s acquisition of WhatsApp, where initial promises of data separation were ultimately broken. In 2017, the European Commission fined Facebook €110 million for providing misleading data regarding the WhatsApp acquisition. Similarly, TikTok recently faced a €530 million fine from the EU for transferring European data to China, a practice the platform initially denied but later acknowledged.

The promises are only engaging those who believe them.
Hélène Lebon, Data Protection Lawyer

Updated Policies and Regulatory Scrutiny

Adding to the controversy, Meta updated its terms of service in late April, stating that public posts and photos on Facebook and Instagram will be used to train its AI. Users who wish to opt out must complete a form by May 27th. This policy change has drawn the attention of European authorities. The CNIL (National Commission for Data Protection) announced in a press release that European bodies are collaborating to assess the compliance of Meta’s data processing practices, specifically focusing on the legality of these practices.

The inherent Challenges of AI Regulation

The rapid advancement of AI technology poses a significant challenge to regulators. As Hélène Lebon notes, GAFAM companies (Google, apple, Facebook/Meta, Amazon, and Microsoft) are constantly experimenting, often outpacing the ability of legislators to keep up. This creates a situation where regulatory action often lags behind technological developments, potentially leading to significant fines years after the initial policy implementations.

The GAFAM are testing on many subjects and the authorities run behind. The technique goes faster than the legislator. In my opinion, it will not be in conformity: they will discuss for months and nothing will happen before they take a big fine, in three-four years.
Hélène Lebon, Data protection Lawyer

User Control and the Uninstallable AI

A key point of contention is the fact that Meta AI cannot be uninstalled from whatsapp, Instagram or Messenger. This lack of user control has raised concerns among policymakers. MEP Veronika Cifrova Ostrihonova has formally addressed the European Commission, highlighting the implications for digital security and user autonomy.

Adding to the controversy,WhatsApp’s director of international interaction suggested that a version of WhatsApp without AI exists,implying that users can simply avoid using the AI features. However,critics argue that this response is overly simplistic and fails to address the underlying concerns about data privacy and user choice.