WhatsApp’s Incognito Chat Makes Meta AI Conversations Private

Meta is trying to make AI chat feel a little less permanent.

The tech giant launched Incognito Chat with Meta AI on May 13, giving WhatsApp users a temporary, text-only mode for private questions that the company cannot read or save. The feature extends WhatsApp’s privacy strategy into generative AI, where users may want help with sensitive financial, health, work, or personal topics without leaving a lasting record.

But the same design that limits Meta’s access to sensitive prompts also raises a harder question: who can scrutinize an AI system when even the company says it cannot see what users asked?

Meta pushes privacy into AI chat

Meta said Incognito Chat is rolling out on WhatsApp and the Meta AI app over the coming months. The company positioned the feature as a private way to ask Meta AI questions without exposing the conversation to WhatsApp or Meta systems. 

“Incognito Chat with Meta AI is truly private, meaning no one, not even Meta, can read your conversations,” Meta emphasized.

When a user starts an Incognito Chat, WhatsApp creates a temporary conversation that disappears by default. The mode is built on its Private Processing technology, which processes requests in a secure environment that Meta cannot access. 

How Private Processing works

According to Meta, Incognito Chat runs on its Private Processing technology, a server-based system that lets AI handle requests without exposing content to Meta or WhatsApp.

“Your messages are processed in a secure environment that even Meta cannot access,” Meta said.

The system uses Trusted Execution Environments, which are isolated computing spaces designed to protect data during processing. Meta’s white paper explained that the approach supports AI tasks too demanding for a phone while limiting access to user messages.

Meta stated Private Processing is optional and user-controlled, meaning the company cannot send data to a system without a user action. The company also plans to expand the same approach to Side Chat with Meta AI, a future feature that will provide private AI help inside a WhatsApp conversation without disrupting the main chat.

Privacy raises accountability questions

Will Cathcart, head of WhatsApp, said that users wanted AI help on sensitive subjects but did not always want to share that information with a company.

“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,” Cathcart said, per the BBC. 

The same privacy design could create accountability questions. Alan Woodward, a cybersecurity expert at the University of Surrey, told the BBC there was a risk of a lack of accountability for the AI’s responses.

“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,” Woodward emphasized. 

Meta said Incognito Chat will initially process text, not images, and that Meta AI guardrails will refuse harmful or illegal requests.

Read more about how Meta’s temporary WhatsApp API access for rival AI chatbots in Europe could affect competition, regulation, and business messaging.