Google announced the launch of Personal Intelligence, a new beta feature in the Gemini app designed to provide users with a highly personalized AI experience by securely connecting information from users’ Google apps, including Gmail, Google Photos, YouTube, and Search. Personal Intelligence enables Gemini to reason across multiple data sources and retrieve specific details to deliver tailored answers and suggestions.
Personal Intelligence is launching as a beta rollout in the United States to eligible Google AI Pro and AI Ultra subscribers. This feature is available on Web, Android, and iOS platforms, and integrates seamlessly with all models in the Gemini model picker. The rollout reflects Google’s commitment to evolving AI from general purpose interaction toward deeply contextual, personalized assistance that reflects users’ real lifestyles and preferences.
Josh Woodward, Vice President of Google Labs, Gemini & AI Studio, has shared internal insights and user examples demonstrating how Personal Intelligence can simplify everyday tasks by connecting app data in context. In one scenario, a user at a tire shop was able to ask Gemini for the correct tire size for a 2019 Honda minivan and receive tailored tire options based on past family road trip photos from Google Photos, along with ratings and pricing. When the user needed their license plate number, Gemini retrieved it from a photo and helped identify vehicle trim using information from Gmail.
Also Read: AccuKnox Announced Partnership with Hexaware Technologies
Google built Personal Intelligence with privacy at the center of the user experience. The feature is off by default and users choose which apps to connect. When enabled, Gemini accesses selected data solely to answer specific requests or perform tasks, and sensitive information remains securely within Google’s infrastructure. Users can adjust connectivity settings anytime, disconnect individual apps, or delete chat history. Gemini provides transparency by attempting to reference or explain which connected sources were leveraged when generating responses. Guardrails are in place to avoid proactive assumptions about sensitive topics, such as health, unless explicitly asked. Gemini does not train directly on private content such as a user’s Gmail inbox or Google Photos library; the model is trained only on limited interactions such as specific prompts and model responses.
Personal Intelligence is now available to eligible subscribers in the United States, with plans for expansion to additional countries and tiers, including the free tier, in the future. It is currently available for personal Google accounts and not for business, enterprise, or education users on Google Workspace.