WhatsApp has rolled out a new "incognito" mode for its AI chatbot, enabling users to engage in conversations that will not be monitored by either the user or the tech company. This feature ensures that past interactions disappear from the user's chat history, offering a layer of privacy previously unavailable.
Will Cathcart, the head of WhatsApp, explained that the company recognized a user desire for private discussions with AI on sensitive topics such as health, relationships, and finances. The incognito mode is designed to address this by preventing these conversations from being stored or accessed. However, a cybersecurity expert has raised concerns that this lack of access could hinder accountability for WhatsApp if issues arise, as there would be no record of the interactions.
Meta, the parent company of WhatsApp, also owns platforms like Instagram, Facebook, and Messenger. The integration of Meta AI into WhatsApp last year faced some user resistance, with some expressing frustration at the inability to disable the feature. Despite these initial reactions, Meta CEO Mark Zuckerberg announced in May that Meta AI had surpassed one billion users across its suite of applications, indicating significant adoption.
Cathcart elaborated on the user feedback, stating, "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." This sentiment highlights a growing demand for AI tools that respect user privacy while still providing valuable information.
Currently, most AI companies retain some user data from chatbot interactions. This data is often used to train and improve future versions of their products, particularly for non-premium users. Zuckerberg has positioned Meta's approach as distinct, describing it as the "first major AI product where there is no log of your conversations stored on servers."
The technology underpinning WhatsApp's incognito mode is distinct from the end-to-end encryption used for standard WhatsApp messages. Nevertheless, Cathcart asserted that it offers an "equivalent" level of privacy for AI interactions.
Professor Alan Woodward, a cybersecurity expert at Surrey University, commented on the potential risks. While he believes the introduction of a separate privacy system poses a low risk to WhatsApp's existing security infrastructure, he expressed concerns about how incognito mode might obscure AI malfunctions or instances of abuse. He noted that several AI companies, including OpenAI and Google, have faced wrongful death lawsuits, underscoring the potential for AI to provide harmful advice.
Woodward highlighted the risk of diminished accountability, stating, "Personally I think what you ask an AI should remain private as some people ask it very personal matters – but you are placing a great deal of trust in the AI not to lead users astray." The core concern is that if disappearing messages cannot be retrieved by either the user or Meta, it would become impossible to establish whether a user's conversations with the AI contributed to harm, death, or suicide.
Initially, the incognito mode will only support text-based interactions, excluding image processing. Meta AI's safety protocols are designed to be cautious, with the AI programmed to refuse requests that could be interpreted as harmful or illegal. WhatsApp has also taken steps to limit the AI chatbots available on its platform, ensuring that users primarily interact with Meta's own AI.
Susannah Streeter from investment platform Wealth Club commented on Meta's substantial investment in artificial intelligence, noting, "Meta is on track to shell out $145bn [£107bn] on AI infrastructure in 2026, and investors want to see a lot more bang for those mega bucks." She added that while investors are apprehensive about the scale of this spending, Meta anticipates significant returns. The company is building AI infrastructure to enhance its platforms, with the potential to further solidify its dominance in advertising and commerce if successful.
The move comes as Meta continues to invest heavily in AI development, a strategy that has recently seen its shares fluctuate as investors weigh the long-term potential against the substantial upfront costs. The company's broader AI initiatives, including Meta AI searches and past investigations into AI interactions, underscore its commitment to integrating artificial intelligence across its services.
