Google's Gemini AI assistant now has Personal Intelligence, a new feature that allows for more context-aware and personalised responses by safely connecting to users' Google apps, including Gmail, Google Photos,
YouTube, and search history.
With the new feature, Gemini can reason across various personal data sources, retrieving particular information and offering customised responses without requiring users to specify where to look.
What Personal Intelligence offers
The feature builds on Gemini's existing reasoning abilities, allowing it to combine information from emails, photos, videos, and past searches. For instance, while shopping for tyres, Gemini reportedly referenced family road trip photos from Google Photos to suggest suitable options and pulled vehicle details like trim and licence plate numbers from Gmail and images.
It also supports practical tasks such as travel planning, purchase decisions, and recommendations for books, shows, clothing, or destinations based on a user's preferences and history. Users can correct preferences (for example, stating a preference for window seats) and regenerate responses without personalisation if desired.
Privacy and user controls remain central
Google emphasised that Personal Intelligence is opt-in and off by default. Users must actively enable it in Gemini settings and choose which apps to connect. Data from connected apps is accessed only to respond to specific queries and is not used to train Gemini models directly—training instead relies on filtered prompts and responses with personal information obfuscated.
The company has implemented guardrails to avoid proactive assumptions on sensitive topics like health, though it will discuss such data if directly asked. Connections can be disconnected at any time, and chat history deleted. Google acknowledged potential limitations in the beta phase, including inaccurate responses, over-personalisation (linking unrelated topics), or challenges with timing and nuance (such as changes in relationships).
Availability and rollout details
The feature launched in beta exclusively in the US for subscribers to Google AI Pro and AI Ultra plans. It is available across the Gemini app on web, Android, and iOS platforms, but only for personal Google accounts -not for Workspace, business, enterprise, or education users.
