WhatsApp has stated a viral social media message claiming WhatsApp group administrators must activate Advanced Chat Privacy to prevent Meta AI from accessing group chats and members’ contacts is false.
The claim had circulated widely among WhatsApp users, warning that group chats could be exposed to “AI-related cybersecurity threats” and that AI agents could legally access private messages and contact details unless administrators switch on the privacy setting.
But WhatsApp says Meta AI cannot access entire conversations or users’ contact lists by default.
What Meta AI can see
Meta AI is an optional artificial intelligence feature integrated into WhatsApp. Users can interact with it directly or invoke it within group chats to ask questions, generate suggestions, create images and perform other tasks.
According to WhatsApp, Meta AI only receives content that users specifically choose to share with it.
For example, a user can invoke Meta AI in a group by mentioning “@Meta AI” or forward a message to the AI. It does not automatically gain access to the entire group conversation.
WhatsApp also says Meta AI cannot access members’ contact details simply because it is available in a group.
“Meta AI on WhatsApp can only read what you choose to share with it (it does not have access to all your chats or contacts),” WhatsApp said in response to the viral claim.
The company added that Meta AI is not active until a user chooses to interact with it, either by messaging it directly or invoking it in an existing conversation.
What Advanced Chat Privacy Does
WhatsApp does have a feature called Advanced Chat Privacy, but its purpose is different from what the viral message suggests.
The setting provides additional controls over what can be taken outside a WhatsApp conversation. When enabled, it can prevent participants from exporting chats, automatically saving media to their devices and using messages with certain AI-powered features.
It therefore provides an additional layer of privacy but is not a switch that prevents Meta AI from secretly reading an entire group chat.
WhatsApp’s existing end-to-end encryption remains the primary protection for personal messages and calls. Under that system, WhatsApp says messages are protected so that only the people involved in the conversation can access their contents.
The misinformation was also addressed by WhatsApp and WhatsApp-focused platform WhatsApp Beta Info.
WhatsApp Beta Info said the viral claim was inaccurate, noting that only messages specifically involving Meta AI are shared with the service. It also said users can check whether Meta AI has read a message through the Message Info screen.
WhatsApp itself directly rejected the claim, saying:
“This is not true. Meta AI on WhatsApp can only read what you choose to share with it.”»
The company reiterated that Meta AI does not have access to all chats or contacts and that users must choose to interact with the feature.
