Google has launched Personal Intelligence for Gemini users in Singapore, enabling the AI assistant to draw on data from connected Google apps to deliver more contextual and personalised responses.

The feature, which first launched in the United States, is now available to Google AI Plus, Pro, and Ultra subscribers in Singapore, with a rollout to free users planned within the coming weeks.

Cross-app context in a single interface

Personal Intelligence allows Gemini to access and reason across user data from apps including Gmail, Google Photos, YouTube, and Search. Users retain full control over which apps are connected, and the feature is switched off by default. When enabled, Gemini can retrieve and combine information across linked sources — for instance, pulling booking confirmation details from Gmail alongside relevant screenshots saved in Google Photos to help a user plan a trip.

The integration works across Gemini on web, Android, and iOS, and is compatible with all models available in the Gemini model picker. At launch, the feature is limited to personal Google accounts and is not available to Workspace business, enterprise, or education users.

Privacy architecture

Google says Personal Intelligence was built with privacy at the centre of its design. User data from connected apps is used only to respond to specific requests and is not used to directly train Gemini on users’ Gmail inboxes or Google Photos libraries. The company notes that because the data already resides within Google’s infrastructure, users do not need to transmit sensitive information to third-party services to enable personalisation.

Users can disconnect apps, delete chat history, and regenerate responses without personalisation at any time. Gemini also aims to avoid making proactive assumptions about sensitive personal data, such as health information, though it will engage with such topics if directly prompted.

Known limitations

Google acknowledges the feature is in beta and may produce inaccurate or over-personalised responses — particularly around timing, nuance, or relationship changes. Users are encouraged to provide corrective feedback directly within the interface.