Google’s Spark AI becomes free Tasks agent

▼ Summary
– Google is rebranding its Spark AI agent as “Tasks” and making the feature free for all users in select countries.
– The new feature is powered by the Gemini 4 Hydrogen AI model and offers capabilities like meal planning, calendar management, and expense tracking.
– US users gain access to advanced functions including automated business calls for price comparison, Google Pay purchases, and account sign-ins via Password Manager.
– Access has expanded from exclusive subscribers to a broader audience across the US, Canada, Japan, Korea, Australia, and India through the Gemini interface.
– Google ensures security by running tasks in hardware-isolated virtual machines and requiring user permission for sensitive actions like financial transactions.
Google has officially rebranded its Spark AI agent as “Tasks” and is rolling it out to the public at no cost. This new capability, driven by Google’s latest Gemini 4 Hydrogen AI model, empowers users to automate complex daily activities. The assistant can generate meal plans, organize calendar schedules, and scan email histories to identify recurring charges that may be unnecessary.
While the core functionality is broad, specific enhancements are currently available to users in the United States. These premium features allow the AI agent to interact with businesses directly by calling to compare product prices and verify inventory levels. Additionally, US residents can instruct the agent to purchase items using Google Pay and log into accounts that support the Google Password Manager. For those preferring text-based interaction, American users can also assign tasks by sending emails to gemini@google.com or via text message.
The tool was initially launched in May as part of the paid Google AI Pro and Ultra subscription tiers. With this update, Google is expanding access to a wider international audience. Users in the US, Canada, Japan, South Korea, Australia, and India can now utilize these agentic capabilities without a subscription fee. To activate the feature, individuals must open Gemini and switch the homepage toggle from “Chat” to “Tasks.”
This release positions Google within a rapidly growing sector of agentic AI competitors, which includes offerings like Claude Duet, OpenAI’s Dots, and Meta’s Muse. Security remains a priority for the company, with Google stating that Gemini executes tasks within a hardware-isolated virtual machine dedicated exclusively to each user account. The system ensures that data is never shared across different user profiles. Furthermore, users retain control over the automation process; they can provide real-time instructions while tasks are running. For high-stakes actions, such as financial transactions, the agent pauses to request explicit user permission before proceeding.
(Source: The Verge)




