Google reorganizes Project Mariner team as AI coding agents boom
Google is reorganizing the team behind Project Mariner, its AI agent for browsing the web. The reason: Silicon Valley has sharply shifted its focus to agents that write code. OpenAI, Anthropic and other labs are placing new bets, and Google is no exception.
AI-processed from Wired; edited by Hamidun News
Google is restructuring the team behind Project Mariner, its Gemini-based agent that can work in a browser on its own and carry out tasks on the web. The reorganization comes as the whole industry sharply shifts its interest toward AI agents for writing code.
Project Mariner appeared in late 2024 as one of the first public demonstrations of a new class of agents: ones able to operate a browser like a human, clicking, filling in forms and following links. At the time, it looked like the next big step. But by 2025–2026, the main battleground had become not web navigation but code generation and editing.
Over the past few months, Silicon Valley has been going through another wave of excitement, this time around agents that can autonomously write, test and deploy software code. OpenAI, Anthropic, Microsoft and a number of startups have put significant resources into exactly this area. In these conditions, browser agents, however technically impressive, have slipped down the list of priorities.
According to Wired, the change in the Project Mariner team's configuration reflects a broader strategic reorientation at Google: resources are being redistributed to where competition is now concentrated. This does not mean the project is being shut down; rather, its emphasis and pace of work are changing.
History is repeating itself: the AI industry moves in waves, and each new hype cycle redistributes talent and funding. Today the flagship is coding agents. What the next breakout segment will be in a year, time will tell, but Google, by all appearances, prefers to keep up with the current trend rather than bet only on long-term technology niches.
- Nov 2, 2026Bun creator uses Claude Code to port 750K lines from Zig to Rust in 11 days
- Nov 2, 2026The Verge reviewer builds a $50 Meta Muse Charm alternative with Claude Code
- Nov 1, 202616 models tested on 7 search tools: Exa averaged 10.5 out of 14, SearXNG about 3
- Nov 1, 2026Google's StreamVibe video agent beats other models by 10.7% on response timing and quality
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.