Meta Says Its AI's New Incognito Mode Is So Private Even Meta Can't See Your Chats

Meta is introducing a new private mode for its AI chatbot and says it's so private that not even the company will be able to see what you talk about. You'll be able to create these private chats in WhatsApp or the standalone Meta AI app as the feature rolls out in the coming months, Meta said Wednesday.

Other chatbots, including Google's Gemini and OpenAI's ChatGPT, have private modes. But those are temporary chats, which disappear after a certain number of hours, and they're still viewable by the company that runs the chatbot. Meta says that this won't be the case for its incognito mode. 

"Your messages are processed in a secure environment that even Meta cannot access. Your conversations are not saved, and by default, your messages disappear -- giving you space to ask questions and explore ideas without anyone watching," the company wrote on Wednesday in a blog post.

Watch this: Latest on the Memory Shortage, New AI Rules for Oscars, and Custom Passes for iOS 27 | Tech Today

03:25

The idea is that with the more private version of Meta AI, you may feel more comfortable asking personal questions. This could be info about your health, finances or career planning. Even if Meta can't see what you're telling the incognito chatbot, you really shouldn't share confidential or sensitive information with AI. These systems aren't perfect, even in incognito mode.

Read more: 8 Essential Security Tips for Using AI Chatbots Safely

It's unclear whether Meta could produce these incognito chats if asked in the course of a lawsuit. We've seen this happen with wrongful death lawsuits against OpenAI and Google, where the victim's chats with AI are central to the case. Chat logs have also come up in copyright lawsuits. In a case between OpenAI and The New York Times, a judge sided with the Times, ordering the ChatGPT maker to produce anonymized logs of user interactions.

(Disclosure: Ziff Davis, CNET's parent company, in 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)