Ask Gemini to handle your multi-step tasks in select Android apps

To use Gemini to complete tasks in Android apps, you need to:

• Have a supported Android device:Pixel 10, Pixel 10 Pro, Pixel 10 Pro XL, or Samsung Galaxy S26 Series.
• Be 18 or over and in the US or Korea.Pixel 10 devices are not supported in Korea.
• Be signed in with a personal Google Account.This feature isn’t available if you sign in to a work, school, or supervised Google Account. Learn how to sign in to Gemini Apps .

Important:

• For now, screen automation is only available in English and for select apps.
• This feature is only available in the Gemini mobile app on Android, including when you activate Gemini by touch or by saying “Hey Google.”
• We’re gradually releasing this feature, so it might not be available to you just yet.

1. On your Android phone, open the Gemini mobile app.
   • You can also activate Gemini by touch or by saying “Hey Google.”
2. Tell Gemini the details of the task you want it to complete in an app, like “Book a ride to the airport.”
   • If Gemini can complete a task using a connected app , it does that instead of using screen automation. If the app isn’t connected, you’ll get the option to connect it. Learn how to use and manage Connected Apps .
   • Gemini may ask for more details before it creates an action plan for your request.
   • Gemini works on the task in the background, even if you close the chat.
3. For the first time Gemini uses screen automation, follow the on-screen instructions.
   • To complete tasks in apps for you, the Google app needs permission to use screen automation.
4. If needed, review and confirm Gemini’s action plan for your task.
   • To make changes to the plan, submit a prompt that explains what you want updated.
5. Follow the instructions in the chat to complete any steps that you need to complete and to complete the final step, like confirming a ride booking or placing an order.
   • If you ask Gemini to prepare an order but you don’t complete it, items may remain in your cart.

• Book a ride:
  • Book a ride to the airport
  • Schedule a ride for tomorrow
• Order food:
  • Reorder my last coffee
  • Order pizza for delivery
• Order groceries:
  • Add milk and eggs to my grocery cart
  • Order groceries to my mom's house

To view and take control of a task:

1. In the chat or live update notification, tap View progress.
   • This opens a view of the on-device virtual phone environment that Gemini is using to complete your task.
2. To take control of the task, tap Take control.
3. Complete your steps.

To stop the task:

• In the chat text box, tap Stop response Stop task, or
• In the live update notification, tap Stop task.

1. On your Android phone, open the Gemini mobile app.
2. At the top, tap your Profile picor initial Settings Screen automation.
3. Select Always allow, Ask every time, or Do not allow.

Important:

• Before you share a chat, review the public link and make sure it doesn’t include any private or sensitive info.
• When you share a chat, a public link is created and anyone with the link can view it. Learn more about sharing chats from Gemini Apps .

You can share chats where Gemini used screen automation to complete tasks in Android apps. Learn how to share your chats .

There are limits on the number of:

• Requests you can make where Gemini uses screen automation each day
• Tasks Gemini can handle at the same time using screen automation

If you’re close to your limit, Gemini lets you know how many requests are left for the month. Learn more about limits in Gemini Apps .

Agentic features, like screen automation in Gemini, are helpful and powerful tools, but it’s important to be aware of possible risks. Remember that Gemini can make mistakes and do unexpected things.

Gemini can make mistakes

Gemini might misunderstand what you want it to do or the app it’s interacting with. This could lead to unintended actions, like:

• Following instructions incorrectly or tapping the wrong button or item
• Completing a purchase or ordering something without your permission
• Adding the wrong quantity of an item to your cart or forgetting to add an item
• Confirming a task is done when it’s not or could have done more
• Asking you to take over when it can normally do the step

Learn how to view, stop, and take control of a task .

About prompt injection

Prompt injection is an attempt to elicit an unintended or harmful response from generative AI tools like screen automation in Gemini. This happens when content, like a website, in-app content, an email, a document, or multimedia, has malicious instructions that might be hidden from you but visible to the AI agent.

If an AI agent reads these instructions while doing a task, it might be tricked into doing unintended things. For example, a malicious instruction could tell the agent to:

• Place an order in the app without your confirmation
• Change your delivery address or other information in an app
• Send communications through chat functions in the app

Learn about how we address these risks in Gemini .

• “Take control” mode:Gemini may pause and ask you to “take control” of the task to complete specific actions, like:
  • Entering specific types of information (such as passwords or payment details)
  • Finalizing transactions
  • Signing in to an app
  • Creating an account
• Requests for user confirmation:Gemini aims to ask for your review and confirmation before it completes certain actions in the app. For example:
  • Sending communications, like a message to your driver
  • Changing details, like your delivery address
• Prohibited task recognition:Gemini has safeguards designed to help it recognize and not act on certain types of requests that could be harmful or are outside of its intended purpose, including tasks that may violate the Generative AI Prohibited Use Policy .
• App & action restrictions:Gemini is designed to restrict its activity to the apps and actions relevant to your task. This helps protect your data on other unrelated apps and helps prevent actions that don't align with your task.

Gemini’s safeguards for screen automation described above don’t guarantee protection against all risks.

Monitor your tasks

Monitoring your tasks is the most important way to protect against risk while Gemini uses screen automation. Review confirmation requests and clarifying questions closely. Use screen automation carefully and safely, and only give Gemini tasks you’re comfortable with. Be extra watchful when:

• Gemini handles an important task
• Gemini does anything where a mistake could be a problem

You’re in control

If Gemini is doing something you don't like, you can view, stop, and take control of a task at any time. Learn how to view, stop, and take control of a task .

• Gemini Agent:While screen automation only supports tasks in select Android mobile apps, you can use Gemini Agent for multi-step tasks in a live web browser if you’re eligible. Learn more about Gemini Agent .
• Connected Apps:Screen automations and Connected Apps have their own settings. Also, screen automation won’t work on apps that are available as Connected Apps. If screen automation is off, you can still use Connected Apps. Learn more about Connected Apps .