WhatsApp Rolls Out ‘Private Processing’ to Deliver AI Features Without Compromising Privacy
WhatsApp has unveiled a new feature called Private Processing, designed to let users access advanced AI tools while ensuring their data remains secure and private.
Built using confidential computing technology and operating within a Trusted Execution Environment (TEE), Private Processing allows users to interact with AI — for tasks like summarizing messages or generating writing suggestions — without Meta or WhatsApp gaining access to those requests. “Our goal is to provide AI functionality with the privacy WhatsApp users expect,” the company stated in a blog post.
Here’s how it works: the user’s WhatsApp client first obtains anonymous credentials to confirm identity. Then, it fetches encrypted keys (HPKE) from a third-party CDN, ensuring that AI requests can’t be linked to individual users.
To further protect anonymity, requests are routed through a third-party relay using an Oblivious HTTP (OHTTP) connection, which conceals the user’s IP address. A secure session is then created between the user’s device and Meta’s TEE via RA-TLS (Remote Attestation + TLS).
Once connected, the AI request is sent encrypted using a temporary key to a confidential virtual machine (CVM) within the TEE. The data is processed securely — no messages are stored — and the response is returned to the user, encrypted with a key known only to the device and the processing server.
Private Processing is entirely optional and will not be enabled by default, allowing users to choose if and when they use it. The feature will roll out gradually in the coming weeks.
This follows WhatsApp’s recent release of Advanced Chat Privacy, which limits actions like chat exports and auto media downloads — offering users even greater control over their conversations.
Found this article interesting? Follow us on X(Twitter) ,Threads and FaceBook to read more exclusive content we post.
