Google closes Project Mariner: browser AI agent integrated into Gemini

Google closed Project Mariner on May 4, 2026 — a browser-based AI agent that could perform online tasks on behalf of the user. A brief message appeared on the project's homepage: "Project Mariner was shut down on May 4, 2026. Its technology continues to work in other Google products." This is not the end of Mariner's story — it is a transition to the next stage of its life within the company's corporate structure.

A Year of Intensive Development: From Idea to Parallel Tasks

Project Mariner debuted in December 2024 as one of Google's most ambitious AI experiments. The concept was simple but innovative: an AI agent sees the screen the way a human does, clicks buttons, fills out web forms, reads results, and reports on completed work. The first months were modest in capabilities — the agent sequentially performed one task at a time, learning to interact with different interfaces. But over a year of development, Google's team significantly expanded Mariner's capabilities.

At the Google I/O 2025 conference, the company presented a major update: the agent could now execute up to 10 tasks in parallel simultaneously. This was a leap in performance. Mariner could open multiple browser tabs, work in different web services in parallel, and synchronize results between them. In public demonstrations, Mariner ordered food through delivery, purchased flight tickets, compared product prices — all in one browser, working the way a human would, but faster and more accurately. Google posted impressive videos. Investors understood: this was not just a lab toy, but a potentially useful tool.

Here is the complete list of capabilities that Mariner accumulated:

• Ordering food and goods through marketplaces, delivery services, and online stores
• Searching for and purchasing tickets for flights, trains, and events
• Comparing prices and product specifications across different stores
• Filling out tax forms and various documents
• Automating routine operations in web applications

Technology

Is Not Closed, It Is Integrated Into the Ecosystem

Google did not simply shut down Mariner as a failed experiment. The company integrated its components into the existing ecosystem. The main beneficiary is Gemini Agent, which now received Mariner's browser capabilities. Other Google AI tools will also receive portions of this technology. This is a typical pattern for Google Labs. Experimental projects rarely die. They either become full-fledged products, migrate to existing platforms, or their technology is distributed throughout the company. Mariner chose the path of integration — this is financially sensible and productively logical. Browser agents are needed across Google's entire product line, and maintaining a separate project is inefficient.

Browser Agents Exit the Laboratory Into the Real World

Mariner was an experiment, but 2026 shows that browser-based AI agents are transitioning from the category of curiosities into real functionality. OpenAI released Operator — a built-in feature for web tasks. Google is integrating Mariner's legacy into Gemini. Other companies are working on similar solutions. This is no longer experimental territory — it is a new market. For users, this means one thing: routine tasks like booking tickets, filling out forms, and searching for information may soon be completely delegated to an AI assistant. A person gives a command, the agent executes it in the browser, returns the result. This will save time and reduce errors.