WhatsApp Launches AI Chats That Leave No Trace
News Desk
Islamabad: WhatsApp has introduced a new “incognito” mode for conversations with its AI chatbot, offering users what the company describes as fully private interactions, a move being hailed by privacy advocates but also raising concerns among cyber security experts over transparency and accountability.
The feature allows users to hold conversations with Meta AI without chat histories being stored or monitored, even by the company itself. Once the mode is activated, previous conversations disappear from the user’s chat history and cannot be retrieved later.
According to Will Cathcart, the company developed the feature in response to growing user discomfort about sharing sensitive information with AI systems.
Cathcart said users increasingly want to discuss personal topics such as health, finances and relationships with AI tools while ensuring those conversations remain inaccessible to companies or third parties.
The announcement marks a major shift in the AI industry, where most chatbot providers typically retain at least some user interactions for system improvement and AI training purposes.
Mark Zuckerberg described the tool as the “first major AI product” where conversations are not stored on company servers.
Unlike WhatsApp’s traditional end-to-end encryption system, the new feature reportedly relies on a separate privacy framework that the company says provides an equivalent level of protection.
However, the launch has triggered debate among cyber security experts and digital rights observers.
Alan Woodward warned that while the feature poses little risk to WhatsApp’s existing security infrastructure, disappearing AI chats could create serious accountability gaps if harmful advice or AI malfunctions occur.
Experts note that several AI companies, including OpenAI and Google, have already faced lawsuits linked to alleged harmful AI interactions.
Read More : https://thepenpk.com/whatsapp-lets-users-turn-any-photo-into-a-sticker-in-one-tap/
Critics argue that if conversations cannot be retrieved by users or by Meta, it may become impossible to investigate cases where AI responses allegedly contributed to harm, self-harm or other dangerous outcomes.
Cathcart said the incognito feature will initially support only text-based conversations and that Meta AI’s safeguards would take a cautious approach by refusing potentially harmful or illegal requests.
The move also strengthens Meta’s control over AI interactions within WhatsApp. The platform currently blocks third-party AI chatbots, meaning users can only interact with Meta’s own AI system.
Analysts say the launch reflects Meta’s broader push to dominate the rapidly expanding AI market.
According to investment analyst Susannah Streeter, Meta is expected to spend around $145 billion on AI infrastructure in 2026 as the company seeks to integrate artificial intelligence more deeply into its social media, advertising and commerce ecosystem.
Meta has increasingly positioned AI as central to the future of its platforms after Zuckerberg revealed earlier this year that Meta AI had surpassed one billion users across the company’s apps, including Facebook, Instagram, Messenger and WhatsApp.
