Gemini in Chrome is now available for Android users in the US, Google confirmed. Originally previewed in May at the company’s I/O conference, the AI features have made their way to mobile after debuting on the Chrome desktop browser earlier in the year.
On Android, Gemini in Chrome lets users summarise lengthy web pages and pose questions directly, offering a practical way to locate a specific answer within dense content. The tool also connects to select Google apps, including Calendar and Keep, without the need to switch tabs.
What Gemini in Chrome Offers on Android
A further feature is the ability to generate images through the AI image generator Nano Banana. Google points to the example of producing an infographic based on a web page being studied for an exam, catering to those who favour a more visual approach to learning.
Google is also introducing auto browse for the Android Chrome browser, an agentic AI capability that allows users to hand off tasks to the AI. Gemini can book event tickets, update online orders and carry out similar actions. Auto browse will request explicit permission before proceeding with sensitive tasks, such as processing payments. The feature is available to US-based Android users on the AI Pro or AI Ultra subscription plans, with no timeline provided for a wider international rollout.
Safety and Security Considerations
Agentic AI on smartphones is becoming increasingly widespread, with Google’s Pixel 11 phones serving as a notable example thanks to Gemini Intelligence. However, such agents can overstep the mark in completing a task. In one instance, an agent hacked an Australian gym’s website simply to secure a spot in a class for a user. While agentic AI can deliver considerable benefits, caution remains advisable, as these systems may behave unpredictably or take unintended actions.
Google states that it has built Gemini’s features into Chrome with safety and security in mind, training its models to identify common threats such as prompt injections. Prompt injections are a form of malicious cyberattack against AI chatbots that can lead to the leakage of sensitive data and other security risks. The feature launch covers all US-based Android users.