New Delhi: Google is now expanding the role of its Gemini AI with the latest feature called Personal Intelligence. The feature is designed to make Gemini’s responses more relevant by adding personal context from select Google applications. With the user’s permission, Gemini can use information from Gmail, Google Photos, YouTube and Search history to help answer questions and support planning. The aim is to help users with everyday tasks such as finding information or planning activities, while still keeping control over how their data is used.
The Personal Intelligence feature lets users link selected Google applications to Gemini. Once it is connected, Gemini can pull the details from emails, photos or past searches to answer specific questions. Google has stated that the feature is turned off by default, and people can choose which applications to connect, disconnect them later or switch off personalisation entirely for chat. Personal Intelligence mainly does two things: it can understand information from different sources, and it can pull specific details from places like emails or photos to answer questions. By working across text, images and videos together, it gives answers that are more relevant to the users.
Gemini does not directly train on users’ Gmail inboxes or photo libraries. It references that data only to respond to the request. Google has also mentioned that training is done on limited information, like prompts and responses, after steps are taken to remove or mask personal details. Gemini tries to show or explain where its answers come from so users can check the information. Users can correct it or ask for more details. They can also choose to turn off personalisation for chat or use temporary chat without personal data. The company also stated that Gemini avoids making assumptions about sensitive topics, such as health, unless the user asks.
Personal Intelligence is rolling out over the next week to eligible Google AI Pro and AI Ultra subscribers in the US. It works across the web, Android and iOS, and supports all the Gemini models in the Gemini model picker. It will also come to AI Mode in Search soon. Google also stated it plans to expand access to more countries and eventually to free users, but there is no timeline yet; the feature is not available for Workspace or education accounts.
