Gemini adds ‘Personal Intelligence’ to enable app-based personalisation

Google has introduced Personal Intelligence for Gemini, a new feature that allows users to personalise the AI assistant by connecting select Google apps. Announced on January 14, 2026, the feature is launching in beta in the US.

With Personal Intelligence, users can link apps such as Gmail, Google Photos, YouTube and Search to Gemini through a single tap. Google says the integration is designed to help Gemini deliver more context-aware and relevant responses by drawing on information already available within a user’s Google ecosystem.

According to the company, the feature enhances Gemini’s ability to reason across multiple sources and retrieve specific details from emails, photos or search history to respond to queries. The system can combine inputs across text, images and video to generate tailored answers for tasks such as recommendations, planning or information lookup.

Google has emphasised that privacy controls are central to the feature. App connections are turned off by default, and users can choose which apps to link, modify settings at any time, or disable personalisation for individual chats. Gemini will also indicate or explain which connected sources were used to generate a response.

The company clarified that while Gemini may reference data from connected apps to answer user requests, it does not directly train its models on personal content from Gmail or Google Photos. Model training uses limited, filtered data such as prompts and responses, with steps taken to remove or obscure personal information.