Google launches Deep Research Max, an AI agent that automates weeks of research work by autonomously browsing the web and analyzing data

Google DeepMind released Deep Research Max, powered by its Gemini 3.1 Pro language model (an AI system trained to understand text and perform reasoning). Unlike chat tools that answer one question at a time, this agent works autonomously — it plans research steps, browses websites, and synthesizes findings without waiting for human instructions between each task.

For the first time, developers can connect Deep Research to specialized data sources like financial feeds (stock prices, earnings reports, market data) through the Model Context Protocol — a standardized way to plug in external tools. This means financial analysts, investors, and researchers can now feed proprietary company data or live market feeds directly into the agent.

Knowledge workers who currently spend days researching market trends, competitor moves, or investment opportunities can now run automated research tasks that synthesize findings across dozens of sources in hours. Consultants, investment firms, and business strategists are the first groups likely to replace manual research workflows with this agent.