Mark Zuckerberg introduced Incognito Chat, a new feature for Meta AI that promises to be the first major artificial intelligence product with total privacy. According to the CEO of Meta, in these chats, no records are stored on the company's servers.
The messages disappear once the user ends the chat session, similar to the incognito modes of other AI assistants. However, Meta emphasizes that its version goes further because it incorporates end-to-end encryption, meaning that no one, not even the company itself, can read the conversations.
“Other apps have launched incognito modes, but they can still see the questions that come in and the answers that go out. Incognito Chat with Meta AI is truly private,” Zuckerberg noted. This feature clearly differentiates it from competitors like Google's Gemini or ChatGPT.
For example, Google retains temporary chat data for up to 72 hours, while in ChatGPT, temporary chats can be stored for up to 30 days. In Claude, incognito chats are kept for at least 30 days.
Privacy at the Center of the Discussion
The announcement comes at a time when the privacy of interactions with AI is under scrutiny. Recent lawsuits in the United States and Canada have highlighted how records of conversations with tools like ChatGPT were used in mass shooting cases. Even The New York Times obtained a court order to have conversations stored indefinitely.
