Google Expands Project Mariner AI Agent for Multitasking Web Browsing

At Google I/O 2025, Google unveiled significant advancements to Project Mariner, its experimental AI agent designed to autonomously browse and interact with websites on behalf of users. This update marks a pivotal shift in how AI can transform web navigation by enabling the agent to handle nearly a dozen tasks simultaneously, a major improvement over its initial single-task design.

Previously, Project Mariner ran within the user’s browser, limiting multitasking capabilities and user productivity. The new iteration operates on cloud-based virtual machines, allowing users to continue other work while the AI agent completes tasks in the background. This cloud migration aligns Project Mariner with similar AI agents from OpenAI and Amazon, enhancing scalability and responsiveness.

U.S. subscribers to Google’s AI Ultra plan, priced at $249.99 per month, currently have access to Project Mariner, with plans to expand availability internationally. Google is also integrating Project Mariner’s capabilities into its Gemini API and Vertex AI platforms, empowering developers to build AI-powered applications that leverage autonomous web browsing and task execution.

The practical implications of Project Mariner are vast. Users can delegate complex tasks such as purchasing event tickets or ordering groceries through conversational AI without manually visiting multiple websites. This agentic approach streamlines user workflows and redefines online interactions, potentially reshaping e-commerce and service industries.

Google’s announcement also introduced “Agent Mode,” an AI experience combining web browsing with research tools and integrations across Google apps, soon accessible to Ultra subscribers on desktop. Additionally, “AI Mode” will debut within Google Search Labs, offering enhanced AI-powered search experiences with support from partners like Ticketmaster and StubHub.

While still experimental, Project Mariner and related agentic experiences signal a fundamental shift in internet usage, where AI agents increasingly act as intermediaries between users and the web. This evolution promises to enhance productivity, personalize user interactions, and open new avenues for businesses to engage customers.

For developers and enterprises, integrating Project Mariner through Gemini API and Vertex AI offers opportunities to innovate AI-driven applications that automate complex workflows and deliver seamless user experiences. As Google continues to refine these AI agents, early adoption and experimentation can provide competitive advantages in the evolving digital landscape.