LONDON — Meta Platforms announced on Wednesday that it is introducing an “incognito” mode for WhatsApp users. This feature is aimed at providing a more private way to engage with its AI chatbot, addressing growing privacy concerns regarding sensitive information users may discuss in chats.
In a blog update, the company explained that the incognito chat mode enables temporary and private conversations with its AI assistant available on WhatsApp. According to Meta, messages in this mode will be handled in a “secure environment,” ensuring that even the company cannot access them. By default, these messages won’t be saved and will vanish after the session ends.
There have been ongoing concerns about the privacy of generative AI systems. They rely on extensive data, which sometimes includes personal information shared by users during interactions with AI chatbots.
Other chatbot developers have already implemented various privacy features. For instance, Google’s Gemini allows users to disable chat history and opt out of data usage for AI model training, while ChatGPT offers similar options.
Meta’s initiative to launch incognito chats stems from users often posing sensitive queries. “As we engage with AI systems, people are asking meaningful questions about their lives, and it feels wrong to have to share that information with the companies behind these systems,” said Will Cathcart, head of WhatsApp.
The incognito mode includes safety mechanisms that prevent the chatbot from addressing harmful topics. “It aims to guide users towards useful information, refuse inappropriate questions, and may ultimately stop interacting if necessary,” Cathcart explained.
Only text-based interactions will be possible; users cannot upload images or generate them. Additionally, an age confirmation will be required, as Meta restricts access for users under 13.
