Google has taken a new step in the evolution of its artificial intelligence with Gemini Spark, the personal agent who works behind the scenes And now it's reaching more users worldwide. The company announced the expansion of the service to more than 160 countries for subscribers of the Google AI Pro and Google AI Ultra plans, although with significant geographical exclusions affecting Europe.
Furthermore, the big news of the last few weeks is the Spark integration with Chrome's auto-navigate featureThis feature allows agents to use saved accounts and passwords to perform complex tasks such as booking flights or searching for apartments. However, it is currently only available in the United States and requires explicit user authorization for sensitive actions like payments.
Gemini Spark's global expansion: available in more than 160 countries
As of July 30, 2026, Google AI Pro subscribers can access Gemini Spark in more than 160 countries, the company confirmed on its official blog. The tool operates on Google's cloud infrastructure. And it can perform tasks even when the computer is turned off or the phone is locked. However, the fine print reveals that Spark is still unavailable in the European Economic Area, the UK, Switzerland, and Nigeria, likely due to regulatory issues such as GDPR and the European AI Act.
In order to use Spark, in addition to the corresponding subscription, You need to have the 'Save activity' function enabled and have a personal Google account (not a work or school account). The company warns that users should be careful with the sensitive information they enter into scheduled tasks.
The paid plans are divided into Google AI Pro and Google AI Ultra. Although Spark was initially only available for Ultra, The expansion to Pro has democratized access In regions like Latin America, where implementation has already begun, Google has highlighted use cases such as automated travel planning, subscription management, and the creation of personalized customer responses.
Automatic navigation in Chrome: the agent that acts for you
One of the most talked-about features is the Chrome's automatic navigation, which allows Spark to move around the web using your saved credentialsWith the user's permission, the agent can log in to websites, fill out forms, and complete multi-step processes. For example, they can research flight options and begin booking, or search for apartments and schedule viewings.
Google claims that the system is designed to withstand 'command injection' attacks, a technique in which malicious instructions are hidden in web pages. Furthermore, in sensitive actions such as payments, Spark returns control to the user. so that he can make the final decision. A bar at the top of Chrome indicates when the agent is active, adding transparency.
This feature has initially launched only in the United States, but Google has promised to expand it to more regions in the future. The integration with Chrome represents a qualitative leap towards the assistant's total autonomysimilar to how Google redesigns search with AI to make it more conversational, although it has also generated debate about security and data centralization, especially in the crypto community, which sees it as an argument for decentralized identity.
What's new in July's 'Gemini Drop': Voice mode on macOS and new models
On July 31, Google published its monthly summary of new features, known as 'Gemini Drop', with seven updates. The most notable feature for desktop users is voice mode in macOS.This feature allows you to dictate text, transform selected content, or generate images directly from any active window, without needing to open the Gemini app. This function is currently only available in English.
In addition, two new models in the Flash family have been launched: Gemini 3.6 Flash and Gemini 3.5 Flash-LiteThese models, which Google claims offer better reasoning and greater speed, are designed for applications where latency is critical. The ability to create a custom avatar to appear in AI-generated images has also been added, along with new integrations with Dropbox, Zillow Rentals, and Viator.
Another important novelty is that Gemini Spark can now react to events in real timesuch as breaking news or sports scores, and notify the user. This capability is in addition to the existing ability to work with local files on macOS, although the latter is still in beta and requires a Google AI Ultra subscription in the United States.
Gemini Spark Use Cases and Operation
Gemini Spark is not your typical conversational assistant. It is an agent that performs tasks autonomously following prior instructions.without requiring user intervention. For example, it can clean the inbox by summarizing emails, unsubscribe from advertising lists, or generate reports from local files.
Practical examples shared by Google include: organize a trip based on confirmation emailsPlan your weekend by searching for local events, manage sports activities by checking the weather, coordinate events by gathering dietary restrictions, audit monthly subscriptions, and write personalized responses to clients. You can also analyze the user's writing style to create a reusable guide.
The integration with Google Workspace is native, which means that Spark can read and modify documents, spreadsheets, and calendars without any additional configuration.In addition, the connection with third-party applications such as Canva, Instacart, OpenTable, Dropbox and Zillow has been expanded, allowing its capabilities to be extended to more services.
To activate Spark, the user must define the instructions once, and the agent executes them repeatedly. The user is always in control.The user can decide which applications to access and must authorize high-impact actions. The tool is available in the Gemini mobile app, the web version, and the native macOS app.
With all these new features, Google is consolidating its commitment to AI agents that work in the background, although geographical fragmentation remains an obstacle for European users. The expansion to more than 160 countries and automatic navigation in Chrome mark a turning point. in the way of interacting with artificial intelligence, but the absence in the EU and the UK leaves millions of people unable to try these functions.
