WhatsApp is taking an important step towards the future of artificial intelligence without compromising privacy, or at least that promises. The finishing messaging application has announced the incorporation of a new technology called “Private Processing”, a system that, as explained by the company itself, is based on the same approach as Apple introduced with “Private Cloud Compute.”
The idea of goal is to offer artificial intelligence functions such as messages of messages or writing suggestions, without the goal can read a single word of your conversations.
Goal AI is responsible with your data
Apple was the first to use this concept with its Apple Intelligence Functionswhere data processing is done, whenever possible, on the user's own device. And if the task requires more power, it derives its PCC (Private Cloud Compute) servers, specifically designed to ensure that personal data is extremely encrypted and erased completely after processing. Nothing is stored. It is as if the information had never existed, and the best of all is that Apple itself allows independent researchers to verify that this is.
Well, now goal has decided to follow that same path. And it is no accident. The company has been dragging a doubtful reputation for years. Therefore, by recently adding an artificial intelligence chatbot in WhatsApp, visible with a Meta AI circular logo On the chats screen or through a renewed search bar, many users reacted suspicion, since it cannot be deleted.
The problem is that for a chatbot to summarize your conversations, you need to read them. And that, in an environment as private as WhatsApp, awakens all alarms. This is where “Private Processing” comes into play, a confidential computing infrastructure based on what Meta calls a “Trusted Execution Environment”, and that replicates Apple's logic: cloud processingbut without a trace. The messages are not stored even temporarily, and once the work is over, the information disappears without the possibility of recovery.
Apple: The first to do it
Meta ensures that this technology meets two key principles: processing without state (Stateless Processing), which prevents access to past data once the session is finished, and progressive security (Forward Security), which prevents any attack can compromise previous conversations. In addition, they have promised that both users and cybersecurity experts will be able to audit the system to verify that the promises are fulfilled.
Despite the skepticism that can generate any privacy announcement by the goal, the exact implementation of the same standards that Apple has developed and published openly is an important gesture.
The time and independent audits will say whether Private Processing is as safe as it appears, but the intention is clear: goal wants to demonstrate that it can play in the same league as Apple when it comes to protecting user data, even while entering artificial intelligence into its services. And although it is hard to believe it, this time it seems that they are doing well. Is it enough for users to lower their guard?
