Meta Introduces Incognito Mode for AI Chats on WhatsApp

Meta has added a new Incognito Chat mode to its Meta AI chatbot on WhatsApp, designed to give users greater privacy when interacting with the AI. The feature allows users to have temporary, private conversations that are not saved and cannot be accessed by Meta, even for improving the model.

According to Meta, the update addresses growing user concerns about sharing sensitive information with AI tools. Many people now use chatbots for personal, financial, health, or work-related questions. In standard AI interactions, user prompts are typically used for training and refinement. With Incognito Chat, conversations are processed in a secure environment that Meta says even the company cannot access, and messages disappear by default.

The company described it as a space for users to “think and explore ideas without anyone watching.” This aligns WhatsApp’s AI features more closely with the app’s longstanding emphasis on end-to-end encryption and privacy.

Meta is also rolling out a “Side Chat” option, which lets users consult the AI within an existing conversation thread without interrupting the main chat.

While the privacy enhancements may appeal to users wary of data usage, the move has raised questions about oversight and responsibility. Critics point out that removing Meta’s ability to review conversations could limit the company’s capacity to catch harmful or inaccurate responses. AI chatbots are known to produce hallucinations and incorrect information, particularly on sensitive topics such as health advice or potentially dangerous queries.