Gemini now use your emails and photos for personalized help

The feature allows Gemini to draw on information from services such as Gmail, Google Photos, YouTube, and Search to provide answers that reflect individual needs and habits.

How it works

Personal Intelligence is optional and turned off by default. Users choose whether to activate it and which apps to connect, all through a simple setup process. Once enabled, Gemini can reason across emails, photos, text, and videos to answer questions or complete tasks more effectively.

For example, Gemini can retrieve specific details from a photo or email while also applying broader reasoning. In one demonstration, the assistant identified vehicle details from stored emails and photos, suggested product options based on past trips, and pulled exact information such as a license plate number when asked.

Google highlights a range of practical applications, from shopping and travel planning to entertainment recommendations. By analyzing past trips, interests, and communications, Gemini can suggest book titles, shows, or vacation ideas that avoid generic tourist options and better match a user’s preferences.

The system can also be corrected in real time. If Gemini makes an incorrect assumption, users can clarify their preferences, regenerate a response without personalization, or start a temporary chat that does not use connected data.

Privacy and user control

Google emphasizes that privacy is central to the design. Personal Intelligence only accesses connected app data to respond to specific requests and does not directly train AI models on the contents of Gmail inboxes or Photo libraries. According to the company, training focuses on filtered and obfuscated prompts and responses rather than raw personal data.

Availability

The rollout begins as a beta for eligible Google AI Pro and AI Ultra subscribers in the United States, with access expanding gradually. Google says broader availability, including additional countries and the free tier, is planned over time.

It works across web, Android, and iOS and supports all models available in the Gemini model picker. The feature is limited to personal Google accounts and is not available for Workspace business, enterprise, or education users at launch.

Users can enable it by opening Gemini settings, selecting Personal Intelligence, and choosing which apps to connect.

Earlier, Qazinform News Agency reported that Apple chooses Gemini for its AI model.