WhatsApp Unveils Incognito AI Chat Mode, Sparking Privacy and Accountability Debate

WhatsApp announced a new “incognito” setting for its AI chatbot, promising that neither the company nor its servers will retain the text exchanged. The feature works alongside the platform’s existing end‑to‑end encryption, which already hides regular messages from Meta. In incognito mode, chats disappear from the user’s device after the session ends, and no logs are stored.

Head of WhatsApp, Will Cathcart, said users want to discuss sensitive topics—health, relationships, finances—without the risk of their data being harvested. “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,” he explained. The mode currently supports only text queries; image analysis is not yet enabled. Meta’s AI guardrails will refuse requests that appear harmful or illegal.

Cyber‑security professor Alan Woodward warned that the lack of any record could hinder investigations into harmful outcomes. “What you ask an AI should remain private, but you are placing a great deal of trust in the AI not to lead users astray,” he said. Without logs, regulators and victims would have no evidence if an AI response contributed to self‑harm or fraud.

The move follows criticism of Meta’s AI rollout last year, when users could not disable the feature. Meta CEO Mark Zuckerberg later claimed the AI product was the first major one with no conversation logs stored on servers. The incognito mode is technically separate from the platform’s encryption but is described as an “equivalent” privacy layer.

Industry analysts note that Meta plans to spend $145 billion on AI infrastructure in 2026, betting that privacy‑focused features will drive adoption across its apps. Investors are watching to see whether the privacy promise translates into higher engagement and ad revenue.

What it means: WhatsApp users can now query AI without fear of corporate surveillance, but the trade‑off is a loss of any safety net that a conversation record might provide. Regulators, consumer groups, and Meta will need to monitor how the feature is used and whether additional safeguards become necessary.

What to watch next: how Meta responds to potential misuse cases and whether governments introduce oversight requirements for AI chat logs on messaging platforms.