Meta launches Incognito Chat to keep AI conversations private
Meta CEO Mark Zuckerberg announced a new Incognito Chat mode for Meta AI that the company says keeps conversations off its servers and pairs that promise with end-to-end encryption. According to Meta, messages in Incognito Chat aren’t saved to users’ chat histories and there is “no log of your conversations stored on servers.”
Meta contrasts the feature with other AI “incognito” modes by highlighting encryption: while some services stop saving history, they can still see inputs and outputs. Meta says Incognito Chat is different because it uses end-to-end encryption, meaning, in Meta’s words, that “no one — not even Meta — can read your conversations.”
The move is a practical privacy win for people who want to use generative AI for sensitive or personal topics without adding data to company logs. Incognito Chat could make it easier for users to ask health, finance, legal or other private questions without storing a persistent record, reducing long-term exposure of sensitive prompts.
Meta hasn’t published full rollout timing or independent audits yet, so privacy advocates and users will be watching for verification and details about key management, recovery, and interoperability. Still, as a user-facing product step, Incognito Chat represents a clear push toward giving people more control over their AI conversations — an important direction for mainstream AI adoption.