Google's new Gemini Intelligence platform was one of the most interesting novelties released at today's Android I/O event.
Gemini Intelligence is an AI layer integrated into the operating system that sees what is happening on the screen and can control all applications found on the phone. This means it can be used to create independently operating agents for various purposes.
A prime example of this could be an agent that, just before waking up, reads all notifications, emails, and instant messages that arrived on the phone during the night - and creates a summary of them, gathering the truly important things from the flood of notifications. The same agent could also be configured to make calendar entries automatically if the incoming messages contained something clearly belonging to the calendar and worth remembering later.
So, with Gemini Intelligence, Google aims to change the way mobile phones are used, where the user regularly performs some repetitive routines between different applications - by automating them into self-operating background agents. What's also interesting about the feature is that the agent can also automatically react to what is visible on the screen, meaning it doesn't need to receive a "trigger" from any notification or event for it to initiate a desired process in the background. Processes can also involve the automatic use of several, even dozens, of different applications by the AI.
Agents can also be instructed to start only when requested. For example, by taking a picture of an item of clothing seen in a store, one can ask Gemini Intelligence to search for that product in different stores and order it for home delivery from the cheapest store.
The first Gemini Intelligence features will become available in the summer for the latest Samsung Galaxy and Pixel phones, and the functionality will also expand to other Android devices later in 2026.
A prime example of this could be an agent that, just before waking up, reads all notifications, emails, and instant messages that arrived on the phone during the night - and creates a summary of them, gathering the truly important things from the flood of notifications. The same agent could also be configured to make calendar entries automatically if the incoming messages contained something clearly belonging to the calendar and worth remembering later.
So, with Gemini Intelligence, Google aims to change the way mobile phones are used, where the user regularly performs some repetitive routines between different applications - by automating them into self-operating background agents. What's also interesting about the feature is that the agent can also automatically react to what is visible on the screen, meaning it doesn't need to receive a "trigger" from any notification or event for it to initiate a desired process in the background. Processes can also involve the automatic use of several, even dozens, of different applications by the AI.
Agents can also be instructed to start only when requested. For example, by taking a picture of an item of clothing seen in a store, one can ask Gemini Intelligence to search for that product in different stores and order it for home delivery from the cheapest store.
The first Gemini Intelligence features will become available in the summer for the latest Samsung Galaxy and Pixel phones, and the functionality will also expand to other Android devices later in 2026.
