Google rolled out Gemini Spark as a built‑in feature of Chrome, turning the browser into a platform where an AI assistant can actually navigate the web on a user’s behalf. The move marks the first time the Gemini agent operates beyond a chat window, performing background tasks while the user continues browsing.
To activate Spark, users need the most recent version of Chrome on Windows or macOS, a personal Google account and a subscription to either Google AI Pro or Google AI Ultra. Once those boxes are checked, the Safe Browsing setting must be set to Standard Protection or Enhanced Protection. From there, clicking the three‑dot menu, selecting Settings, then AI Innovations (or Gemini) and enabling the “Let Gemini browse for you” permission brings the feature online.
After enabling, an Ask Gemini panel appears at the top of the browser. When a task is entered, Spark first displays a proposed plan, then asks the user to confirm before proceeding. The first connection request prompts Chrome to ask whether to link the browser to Spark; granting permission unlocks the assistant’s full capabilities.
In a practical test, the author asked Spark to locate the best price for a 65‑inch television across Amazon, Best Buy, Costco and Walmart, apply any promo codes, calculate the final cost and add the item to a cart. Spark navigated each site, compared offers, and paused at the checkout stage, waiting for the user to approve any payment step.
A second experiment combined several errands: planning a family day for two adults and two toddlers. Spark assembled opening hours, travel times, a logical itinerary, secured a restaurant reservation with online booking and prepared museum tickets—all without the user having to switch between Maps, review sites and booking pages. As with the shopping task, Spark stopped before finalizing any reservation, seeking explicit approval.
The assistant also proved handy for paperwork. When asked to fill out a library‑card application for a child, Spark pulled address and contact details from the user’s Google profile, populated the form on the library’s website, and halted at the submission button, letting the user review the entries first. Google suggests the same approach could streamline more complex forms such as medical or insurance paperwork.
Google positions Spark as a tool for the “tedious, repetitive tasks” that dominate everyday browsing. By handling tab juggling, data entry and multi‑site research, the AI aims to reduce the manual clicking and typing that often slow users down. While Spark cannot complete purchases or bookings without user consent, its ability to act autonomously within the browser represents a notable step forward for consumer‑focused AI integration.
Este artigo foi escrito com a assistência de IA.
News Factory APP - notícias agênticas para impulsionar seu SEO e AEO.