WhatsApp for Android might be getting a new feature for its artificial intelligence (AI) chatbot Meta AI. As per a feature tracker, the instant messaging platform is adding a new chat memory feature that will let users ask the AI to remember specific information about themselves. Once a piece of information has been saved in the chatbot’s memory, it is said to customise its responses then keeping it in mind, allowing users to get a more personalised experience. The feature was spotted in the app’s latest beta version, but is currently unavailable to beta testers.

According to the WhatsApp feature tracker WABetaInfo, the new feature will make Meta AI more personalised to users. The new feature was reportedly spotted in the WhatsApp for Android beta version 2.24.22.9. However, the feature is not visible currently and those who have signed up for the Google Beta Programme will not be able to try the feature out.

WhatsApp for Android chat memory feature for Meta AI
Photo Credit: WABetaInfo

Dubbed chat memory, the new feature allows users to ask Meta AI to remember specific information about themselves. The feature tracker also shared a couple of screenshots of the feature which highlights how it might work. Based on the images, the option for Memory will be added to the profile page of Meta AI. Users can access it by tapping the Meta AI title within the chat interface.

Once on the profile page, users can see this new option which is described as “What Meta AI remembers about you”. This menu option can be accessed and whenever a piece of information has been saved, it will appear there.

The feature tracker claims users can ask the AI to remember information such as dietary preference, profession, birthdays, preferred conversation style, and more. Once saved, the chatbot is said to be able to remember and access this information and answer queries keeping it in mind. For instance, a vegan user might get only vegan food recipes when they ask for suggestions.

Since the feature is currently in development, details around the feature’s privacy and security are currently not known. Notably, OpenAI’s ChatGPT also has a similar feature, where users have the option to choose which information is saved by the AI and which is not. Users can also manually remove a piece of information from memory if they feel the AI should not have access to it. OpenAI claims none of the information saved in AI’s memory is used to train any large language models.