Google I/O 2026— Only a few days are left for the start of Google’s annual developer conference Google I/O 2026. In this event, the company is going to make many big announcements in the field of Artificial Intelligence (AI). But just before this mega event, a big leak has come to light in the tech world. According to information leaked by a user on Twitter, Google is working on a very advanced AI agent, which has been named ‘Gemini Spark’. This new assistant will be completely different from existing AI tools, because it will not need any manual command or prompt from you to work.

Google I/O 2026: Google is bringing ‘Gemini Spark’ AI agent that works without prompts, big revelation from the leak

Gemini Spark will work in background without prompt

According to leaked reports, Gemini Spark is in the pipeline as an experimental feature. At present, to activate chatbots or AI assistants, the user has to type or give voice commands. In contrast, Gemini Spark will not wait for a user prompt. It will handle the work completely proactively. This simply means that it will be continuously active in the background of your phone, will understand your patterns and will take decisions on your behalf and complete the tasks.

Will complete the task itself by accessing data from linked apps

This new AI agent will be deeply linked with other apps present in the user’s smartphone. According to the leak, it has the ability to access data from your mail, messages, calendar and other third-party apps. For example, if you get an email from a flight ticket or a meeting, Gemini Spark will schedule the event in the calendar without you asking, send you important reminders, and can even process tasks like booking a cab in the background if needed.

Opinion of tech experts and leakers

What will be its impact on common users and next step?

Google can present this feature on the stage of Google I/O 2026 to be held next week. Initially, it will be introduced as an experimental tool, the trial of which is expected to start with selected developers and beta testers. This feature will be rolled out for common smartphone users only after passing stringent standards of privacy and security. If this leak proves to be true, then in the coming days the need for users to type repeatedly on the mobile screen will reduce significantly.