WhatsApp launches 'incognito' AI chat with private disappearing messages

WhatsApp has launched an 'incognito' mode for its AI chatbot, making conversations unreadable by the company and causing past chats to disappear ^[1][2]. The feature is designed for sensitive topics like health and finances ^[2].

The new mode, announced by WhatsApp head Will Cathcart, aims to allow users to have private conversations on sensitive subjects without the company monitoring the exchanges ^[2]. "We've heard from a lot of people that they feel some discomfort about sharing [personal] information with the company, yet they want the answers," Cathcart said ^[2]. Meta CEO Mark Zuckerberg described it as the "first major AI product where there is no log of your conversations stored on servers" ^[2].

However, cybersecurity experts have raised concerns. Professor Alan Woodward of Surrey University warned the BBC that the feature could lead to a lack of accountability for the AI's responses, as neither the user nor Meta would have access to the chat history if something went wrong ^[2]. The concern is that disappearing messages could make it impossible to find evidence if a chat led to harm ^[2].

Cathcart stated that incognito mode will initially only process text rather than images, and Meta AI's safety systems will err on the side of caution in refusing harmful requests ^[2]. The announcement comes as Meta's AI push continues to scale. Zuckerberg said in May 2025 that Meta AI had reached a billion users across its apps ^[1], and the company plans to spend $145bn (£107bn) on AI infrastructure in 2026 ^[1][2].