Gemini is becoming more personal in its responses thanks to Personal Intelligence, a feature that responds to personal preferences and data.
Google today announced Personal Intelligence. This is a new feature that allows users to have the Gemini chatbot respond to their personal preferences and data from other Google services.
Access to Gmail, Photos and YouTube
With Personal Intelligence, Google’s Gemini chatbot gains access to selected user data from Gmail, Google Photos, and YouTube, among other sources. Based on this, Gemini can refine its answers. For example, when asked about recipes, the chatbot can take into account restaurant reservations in the inbox, or use photos of a car to provide maintenance advice.
The feature runs on Gemini 3, which can handle prompts up to 1 million tokens. Because personal data sources are often larger, Google uses a technique called context packing, which selects only relevant fragments from datasets.
Transparency and privacy controls
Gemini shows which data sources are used in an answer, so users can check where the information comes from. According to Google, sensitive data, such as health information, is by default used in a limited way and only included if the user explicitly requests it. Personal Intelligence is disabled by default and users decide which services are accessible. With temporary chats, personalization can be disabled per conversation.
Personal Intelligence will first be available to a limited group of paying users in the US with a Google AI Pro or Ultra subscription. Free accounts, international markets, and Google Search’s AI Mode will follow later.
read also