Google has announced the latest advancements in Project Mariner, a next-generation AI agent that can perform automation tasks and navigate the web like a digital executive assistant. Integrated into the Gemini app’s new Agent Mode, Project Mariner leverages advanced vision-language models to interpret web content and take actions on behalf of users.
The AI agent aims to act directly on behalf of the user by understanding and manipulating web content, such as clicking, scrolling, and submitting forms. It can also replicate complex sequences across similar contexts with its “Teach and Repeat” capability. In this latest release, Mariner can handle up to 10 simultaneous tasks, making it capable of autonomously comparing and booking travel options or filtering data across web apps.
Google’s Agent Mode in Gemini allows users to delegate tasks without needing to click. The AI not only retrieves information via links but also performs digital work on behalf of the user. Currently available in an experimental phase, Project Mariner is expected to be accessible through Google’s AI Mode in Search, initially via the Search Labs program.
Mariner fits squarely in the Agentic AI camp, which implies a more cognitive, self-reflective system that adapts and plans dynamically without hardcoded steps. This approach marks a departure from traditional automation approaches, such as platforms like Make.com and n8n, which require technical fluency and upfront effort.
The emergence of Project Mariner highlights the growing importance of AI-powered automation in digital ecosystems. With its multimodal capabilities, agility, and sustainability, Google’s next-gen AI agent is poised to revolutionize how we interact with technology.
Source: https://www.forbes.com/sites/ronschmelzer/2025/05/20/google-geminis-agent-mode-and-project-mariner-shows-the-future-of-ai-agents