Meta Adds Privacy Layer to WhatsApp AI With Incognito Chat
Meta is launching "Incognito Chat" for its AI assistant on WhatsApp, a feature designed to keep conversations private from the company and anyone else who might access them. Messages sent in incognito mode won't be saved and will disappear by default.
The feature uses Meta's private processing technology, meaning the company cannot access or store the conversations. This addresses a practical problem: people routinely ask AI assistants sensitive questions about finances, health, work, and personal matters while worried about how companies might use that data.
"We're starting to ask a lot of meaningful questions about our lives with AI systems," said Will Cathcart, head of WhatsApp, in a media briefing. "It doesn't always feel like you should have to share the information behind those questions with the companies that run those AI systems."
What This Means for Data Practices
Meta's standard policy allows the company to use messages shared with Meta AI to improve its AI models. WhatsApp's regular chats have long been protected by end-to-end encryption, but those protections don't extend to AI interactions. Incognito Chat closes that gap.
For now, the feature supports text only. Users cannot upload images in incognito conversations.
Relevance for Communications Teams
This move signals how major platforms are responding to privacy expectations. PR and communications professionals should understand how their organizations handle AI-generated content and user data, particularly as employees use AI assistants for work-related tasks on messaging platforms.
The announcement also reflects broader industry pressure to separate AI functionality from data collection practices-a distinction worth monitoring as other platforms introduce similar features.
Your membership also unlocks:
