Gemini takes a decisive step towards truly personal artificial intelligence with the launch of Personal Intelligence, a new capability that securely connects Google apps in a single tap. Rolling out as a beta to eligible users in the United States, this update makes Gemini more personal, more proactive and more powerful than ever before.

Personal Intelligence allows users to connect apps such as Gmail, Google Photos, YouTube and Search to Gemini, transforming it into an assistant that understands context, preferences and real life needs. The feature is optional, off by default, and fully controlled by the user, ensuring personalisation never comes at the cost of trust.

From General Answers To Personal Understanding

While many AI assistants can retrieve information from the web, Personal Intelligence goes further by reasoning across a user’s own content. Gemini can now combine signals from emails, photos, videos and search history to deliver responses that are tailored, timely and genuinely useful.

Whether recalling details buried in an email, identifying information from a photo or suggesting options based on past behaviour, Gemini brings these sources together seamlessly. The result is an assistant that does not just answer questions, but understands situations.

For example, when asked about vehicle details while standing at a service counter, Gemini can identify specifications, recommend suitable options based on past travel patterns, retrieve ratings and prices and even recall licence plate information from a saved photo. The experience feels natural, immediate and grounded in the user’s real world context.

Beyond daily tasks, users are also seeing value in recommendations for books, entertainment, clothing and travel. By learning from interests and past experiences, Gemini can skip generic suggestions and offer ideas that feel thoughtfully chosen.

Built With Privacy At The Centre

Personal Intelligence has been designed with privacy as a foundational principle. Users choose if and when to connect apps, select exactly which services to enable and can disconnect them at any time. When activated, Gemini only accesses data to respond to specific requests or to complete tasks on the user’s behalf.

Crucially, personal content from Gmail and Google Photos is not used to directly train Gemini. Instead, data is referenced moment by moment to deliver an answer. Model improvements rely on limited information such as prompts and responses, with personal details filtered or obscured.

Transparency is built into the experience. Gemini aims to explain where information comes from so users can verify responses, request clarification or correct assumptions in real time. For sensitive topics, Gemini avoids proactive inferences while remaining responsive when users choose to engage.

Learning Together Through Beta

As a beta release, Personal Intelligence continues to evolve. While extensive testing has reduced errors, users may still encounter inaccuracies or moments of over personalisation. Feedback plays a vital role, and users are encouraged to rate responses to help refine the system.

Gemini is also learning to better understand nuance over time. Interests, relationships and preferences can change, and users remain in control by simply correcting the assistant when needed.

Availability And Next Steps

Access to Personal Intelligence with Connected Apps begins rolling out this week to eligible Google AI Pro and AI Ultra subscribers in the United States. Once enabled, it works across web, Android and iOS, and across all Gemini models. Expansion to additional countries and broader user tiers is planned, with availability for personal Google accounts only during this phase.

To enable the feature, users can open Gemini, go to Settings, select Personal Intelligence and choose the apps they wish to connect.

With Personal Intelligence, Gemini moves closer to its ambition of being a truly helpful assistant that understands not just the world, but the individual navigating it.