WhatsApp Introduces "Incognito" Mode for AI Chats, Claimed to Be Safer than ChatGPT

Meta is introducing a new feature called “Incognito Chat” in WhatsApp for conversations with the Meta AI chatbot. This private chat feature with AI is designed to address user concerns regarding privacy when using generative AI chatbots. This is because many users often share sensitive information such as health, finances, and personal matters with AI. WhatsApp’s Incognito Chat is built on a “Private Processing” system, which is even claimed to be inaccessible to Meta itself. The feature was announced by Meta on Wednesday (14/5/2026) and will be rolled out gradually in WhatsApp and the Meta AI app over the coming months. “Incognito Chat with Meta AI is completely private, meaning no one, not even Meta, can read your conversations,” Meta stated in its official announcement. To access this mode, users will simply need to open a conversation with Meta AI in WhatsApp and select the “Incognito Chat” option that appears on the chat page. Once activated, the conversation session will run in private mode and automatically end when the app is closed, the phone is locked, or the user exits the chat. This feature is also said to differ from similar modes in other AI chatbots like ChatGPT, Google Gemini, or Claude, which still store conversations temporarily for a certain period. Meta has also added several security restrictions to prevent the chatbot from providing responses to harmful or inappropriate requests. Additionally, users must be at least 13 years old to use this feature. In the future, Meta plans to introduce a “Side Chat” feature, an AI that can assist users in the middle of WhatsApp conversations without disrupting the main chat.