Meta Launches 'Incognito' Mode for WhatsApp AI Chats, Enhancing User Privacy

Meta Platforms has announced an 'incognito' mode for WhatsApp users, focusing on privacy in AI-assisted conversations. This feature allows private chats with Meta AI, where messages are secure and temporary. The launch addresses privacy issues linked to large language models trained on vast data sets.

Meta Platforms has unveiled a new 'incognito' mode for users of its WhatsApp messenger, aimed at ensuring privacy in conversations with its AI chatbot. The mode allows users to engage in private and temporary chats where messages are processed securely, aren’t saved by default, and disappear upon exiting a session.

This move by the social media giant comes amid increasing concerns surrounding generative AI systems and the privacy of user data. Large language models, which power these AI systems, are often trained on extensive data that can include personal user information shared in AI chats.

Rival AI platforms like Google's Gemini and ChatGPT already offer some privacy features. Meta's new incognito mode addresses the common practice of users asking sensitive questions involving personal, financial, or health data. This mode includes safeguards to steer users away from harmful topics, maintaining a focus on helpful interactions.