Gemini Spark Expands to 160+ Countries with Chrome Integration on the Way

Google announced on July 30, 2026 that Gemini Spark, its personal AI agent first shown at Google I/O 2026 in May, is now available in more than 160 countries. Japan was included, with a Japanese-language blog post going up on July 29. Access requires a Google AI Ultra or Google AI Pro subscription. At the same time, Google revealed a new Chrome browser integration that lets Spark act on the web using your saved credentials, though that feature starts in the US only.

What happened

Detail	Fact
Announcement date	July 30, 2026
Countries reached	160+
Japan availability	Confirmed July 29 (Japanese blog); live for Pro users as of August 2
Supported languages (Japan)	Japanese and English
Required plan	Google AI Ultra or Google AI Pro
AI model inside	Gemini 3.5
Chrome integration	US first, other regions later

Gemini Spark was originally announced at Google I/O 2026 in May. It is a proactive AI agent, meaning it can wake itself up in the background and take action without you opening an app. It runs on Google’s cloud infrastructure, so no local setup is required and it keeps working even when your PC is closed or your phone is locked.

Triggers for that background activity include things like receiving an email on a specific topic or getting a chat notification. When triggered, Spark can process a task ahead of time and have a result ready for you.

What Spark can connect to

Out of the box, Spark integrates with Gmail, Google Docs, and Google Sheets to produce output that reflects your personal context, not just generic answers. A June update added connections to Canva, Dropbox, Instacart, OpenTable, and Zillow Rentals, plus support for custom MCP server connections (a standard protocol that lets AI agents talk to external tools and services).

Two practical examples Google gives:

• Travel planning: Automatically add booking confirmation details to a Google Sheets itinerary every time a new reservation email arrives, then find local events for free days in the schedule and add them to your calendar.
• Subscription management: On the first of each month, scan the previous month’s digital invoices in Gmail, flag price increases or trials about to expire, and draft cancellation emails in advance.

How does the Chrome integration work?

Previously, when Spark needed to browse the web it used a remote cloud-based browser and stored any login credentials on that remote browser. The new Chrome integration changes that. If a user grants permission, Spark can use the accounts you are already signed into in Chrome and the passwords Chrome has saved to complete tasks directly in your local browser.

Google describes use cases such as booking a viewing for a saved rental listing, or researching flight options and starting the checkout process. Before the feature activates, it asks for explicit user permission. Google states it is building in protection against prompt injection attacks (a technique where malicious web content tries to hijack an AI agent’s instructions). For high-stakes actions like payments, Spark hands control back to the user rather than completing them automatically.

Who can access Spark right now?

Google AI Ultra subscribers can use Spark today in the supported countries. Google AI Pro subscribers in Japan are told to expect access “within the next few weeks.” Once enabled, a toggle appears at the top of the Gemini sidebar, letting you switch between standard chat and Spark (currently in beta). Users choose which apps Spark can connect to and can turn the feature off at any time.

Why it matters

Proactive AI agents that act without a prompt represent a meaningful shift from chatbots. Instead of you asking a question, the agent watches for conditions and acts on your behalf. For business operators, that opens up genuine automation of repetitive inbox work, but it also means a paid subscription to an AI service is no longer just a chat window. It becomes infrastructure running in the background of your accounts.

The Chrome integration is the more consequential piece long-term. Giving an AI agent access to your saved passwords and logged-in sessions is a significant trust decision, and it moves Spark from cloud-only tasks into the full web. For anyone already thinking about AI integration in their business workflows, this is worth watching closely, because it sets a template other platforms will copy.

For context on how quickly the competitive AI landscape is shifting, our earlier piece on how Claude flipped the AI leaderboard shows that Google is not the only one racing to ship agentic features.

Our take

Gemini Spark is not trying to replace a dedicated automation tool. It is aiming at the gap between “I know I should automate this” and “I have actually set it up.” For a small business owner who lives in Gmail and Google Sheets, that gap is real and the travel itinerary and subscription audit examples are genuinely useful, not contrived.

The Chrome integration is where we would urge caution. Storing credentials in a remote browser was already a privacy trade-off. Connecting your local Chrome session, including saved passwords, to an AI agent is a bigger one. Google’s stated safeguards (user permission, human handoff for payments, prompt injection protection) are the right commitments, but the details of how those controls work in practice are not yet public. Watch the rollout closely before enabling it on accounts that touch business finances or client data.

If your team is evaluating how tools like Spark could slot into existing workflows, our workflow automation service covers that assessment practically, not just theoretically.

Source: Bing News · Google Gemini