Google Brings Deep Personalization to Gemini With Personal Intelligence

Google is rolling out a new feature for Gemini that aims to make the AI assistant far more person and useful in everyday situations. It’s called Personal Intelligence, and it allows Gemini to use information from your own Google apps, such as Gmail, Photos, Calendar, Drive, Search history, and YouTube, to deliver answers that are based on your real activity instead of generic web results.

Personal Intelligence lets Gemini understand you better across apps

The big change here is that Gemini can now reason across multiple apps at once. Earlier, the assistant could fetch individual items like a single email or photo if you asked for it. With Personal Intelligence enabled, Gemini can combine details from different sources to better understand context and give more relevant suggestions.

Google says this is powered by its latest Gemini 3 models and a new system designed to solve what it calls the “context packing” problem. Since no AI model can read your entire inbox or photo library at once, Google’s system selects only the most relevant information and sends it to Gemini when needed. This lets the assistant focus on what matters for each question without loading everything at once.

In real-world use, this means Gemini can answer questions using personal details it finds across apps. For example, it could identify your car model from an email, confirm details from a receipt, look at past travel photos, and then suggest products or plans that match your habits. It also works across text, images, and videos. This allows Gemini to pull information from photos and combine it with email or search data in one response.

You can choose which apps Gemini can access

Personal Intelligence is turned off by default, and users must opt in. You can also choose which apps Gemini can access and disconnect them at any time. Google says Gemini does not train on personal data from Gmail or Photos. Those sources are only used to answer specific questions, according to Google.

Google admits Gemini may sometimes make wrong assumptions, mix up timelines, or over-personalize by connecting unrelated details. Well, that’s expected since it’s still in beta. It’s rolling out in beta starting today for Google AI Pro and AI Ultra subscribers in the US. It’s available on Android, iOS, and web. Google says support for more countries, free users, and Search’s AI Mode will arrive later.