EU pressures Meta to open WhatsApp to rival AI chatbots

False chain messages are causing alarm among WhatsApp users, falsely claiming that Meta AI can access and use all chats for training purposes. These messages urge users to change privacy settings, warning that the AI can read every message. However, these claims are unfounded. WhatsApp messages are protected by end-to-end encryption, ensuring that only the sender and recipient can read them, not even Meta itself.

Meta AI is integrated into WhatsApp for features like summarizing missed messages and chat histories. This function, called "Private Processing," is designed to be visible only to the user, with Meta assuring that it cannot see if the AI has been used for this purpose. While Meta did offer EU users a way to opt out of using public data from Facebook and Instagram for AI training, WhatsApp's encrypted nature means its content was never part of that discussion.

“

Messages are end-to-end encrypted. This means that no one can read them, not even Meta.

Concerns about data privacy persist, as users worry about how their personal information might be used in the future, particularly with Meta's long-term plans for "AI companions." The company's core business model, advertising, could also benefit from more personalized ads derived from this data. Meanwhile, the EU's Digital Markets Act is pushing Meta to open up WhatsApp to other AI services. Meta has a short deadline to allow competing AI chatbots, such as ChatGPT and Claude, access to the platform, as the EU Commission fears a "monopoly" situation.

“

Private Processing. Meta promises that this is really only visible to the users and even Meta cannot see if the AI was used for this.