Google's AI Agents Now Fuse Web and Enterprise Data

Google announced major enhancements to its autonomous research agents with the introduction of Deep Research and Deep Research Max. These new agents, powered by the Gemini 3.1 Pro model, enable developers to integrate open web data with private enterprise information via a unified API. They also offer the capability to generate native charts and infographics directly within research reports.

The agents feature support for the Model Context Protocol (MCP), an emerging standard for connecting AI models to external data. This allows Deep Research agents to securely query private databases and third-party data services without compromising sensitive information. Google is collaborating with industry leaders like FactSet and S&P on MCP server designs to facilitate deeper integration with financial data providers.

Deep Research is optimized for speed and interactive use, reducing latency and cost. Deep Research Max utilizes extended test-time compute for more exhaustive analysis, making it suitable for asynchronous background tasks. Both agents are available in public preview through paid tiers of the Gemini API, marking a significant step in enterprise AI research capabilities.