WhatsApp is rolling out a new AI-powered feature designed to help users craft their messages, while also introducing enhanced privacy settings to give users more control over the AI’s presence in their chats.
AI-Powered ‘Writing Help’ Arrives
The new function, named ‘Writing Help’, will offer users AI-generated message suggestions in a variety of styles, including professional, humorous, or supportive tones. Accessible via a new pencil icon within one-to-one or group chats, the feature allows users to select a style and generate text, which can then be further edited before sending.
This update is initially being made available to English-speaking users in the United States. A spokesperson for WhatsApp confirmed plans to expand the feature to other languages and countries by the end of the year. However, its introduction has raised questions about its practical use in the fast-paced environment of instant messaging, with some suggesting it may be more beneficial for longer-form writing than for quick replies.
New Safeguards to Address Privacy Concerns
A key concern for users is privacy. While WhatsApp messages are protected by end-to-end encryption, AI queries are typically processed on cloud data centres. In response, parent company Meta has stated it is employing “private processing technology,” similar to Apple’s Private Cloud Compute. This method is designed to prevent Meta’s AI servers from being able to read the content of the AI-generated suggestions.
The broader Meta AI also runs in the background of all private and group chats, but it can only access messages when a user deliberately activates it by typing “@Meta AI”. In such cases, only the specific message containing the tag is transmitted to the AI, while all other messages in the conversation remain fully encrypted and private to the participants.
‘Enhanced Chat Privacy’ Puts Users in Control
Despite these assurances, the constant background presence of Meta AI has caused unease for some users, particularly in group conversations where there is a risk of the bot being triggered accidentally.
To address this, Meta introduced a new privacy function in April that allows users to banish the AI from their chats. A new toggle switch, labelled “enhanced chat privacy,” is now available in both private and group chat settings. This feature is switched off by default, but when activated, it completely blocks all participants in the chat from summoning Meta AI using the “@” symbol. This ensures the AI cannot be addressed and cannot view any chat content.
Crucially, in group chats, any member can activate or deactivate this additional privacy layer, even without administrator privileges. When the setting is changed, other group members receive a brief notification within the chat informing them of the adjustment.
