With Google joining Claude, Microsoft Copilot, Perplexity, and leading legal AI platforms like Harvey and Legora, NetDocuments is giving customers an open, governed way to bring trusted legal context to the AI tools they choose.

The future of legal AI won’t be defined by a single model, application, or provider. Firms are already building AI strategies that span multiple tools, and they need the freedom to choose the experiences that work best for their lawyers and their business.

That’s why NetDocuments has an open, AI-tool-agnostic approach to Model Context Protocol (MCP): one that allows our customers to bring the governed legal content in NetDocuments to the AI tools they choose.

This week, Google Gemini Enterprise became the latest frontier AI provider to join the NetDocuments MCP ecosystem, alongside Claude, Microsoft Copilot, and Perplexity, as well as legal AI partners, including Harvey and Legora.

Together with our partners, we’re making it easier for firms to use the AI tools they choose with the trusted legal context already in NetDocuments. And because that context is built once and shared across every connection, each new tool and user  a firm adds makes the same investment worth more. 

For the CIO managing vendor sprawl, the Chief Innovation Officer building an AI strategy, and the managing partner looking to maximize ROI from AI investments, the opportunity is bigger than connecting another AI tool. It is creating a context layer that makes every connected AI tool more valuable.

One governed context layer, many AI experiences

The Model Context Protocol (MCP) has become the standard way AI tools connect to enterprise systems. But as we wrote in June, MCP is standard; what an AI tool receives through it is not.

Most connections can give an AI access to search results and document text. NetDocuments can provide something richer: structured, matter-specific legal context that helps an AI understand the relationships across the work rather than forcing it to reconstruct those relationships from documents every time.

Every new AI tool an organization adopts typically means a new integration project: new security review, new data governance assessment, new vendor contract. Multiply that by five or six AI tools and the overhead and risk becomes significant. MCP collapses that to a single governed connection. New tools plug into the same context layer without a new build each time.

NetDocuments MCP capabilities are already making a difference:

We connected Perplexity to our NetDocuments repository through MCP, and the shift was immediate. No uploading documents. No selecting files. Our lawyers simply point AI at a matter, and it works from everything they already have permission to access. The firm’s entire governed document history is available to any AI tool we choose, with permissions already in place. That changes how we think about legal AI entirely.”

Thomas Kline
Director of information Technology
Stark & Stark, comments

For a deeper look at why this architectural difference matters, and what it means for governance and accuracy, read our earlier analysis.

How The Legal Context Graph Makes MCP More Valuable

Last week, we published the Legal Context Engineering Benchmark Report, a new framework for measuring what better context is worth to legal AI — not just in answer quality, but in the cost of producing a correct answer.

The results make the case for context clear. In our benchmark, the same frontier AI model using the NetDocuments Legal Context Graph cut the cost per correct answer by 48%, while maintaining the same level of accuracy.

That matters well beyond any single AI application.

The Legal Context Engineering Benchmark was designed around a simple principle: hold the AI constant, change the context, and measure what happens.

Put simply: when AI starts with better context, it doesn’t have to work as hard to understand the matter. It can spend less time searching through documents, piecing together relationships, and reconstructing information the firm already knows.

That efficiency compounds. The context is built once, and every connected tool draws on the same layer — so the return grows with every additional tool, user, and workload rather than starting over each time. That matters more each quarter, particularly as both frontier models and legal AI applications move toward consumption-based pricing.

For managing partners and firm leaders evaluating AI investments, that’s the larger ROI story. The value of better context isn’t limited to making one AI application more efficient. It’s a foundation that can improve the economics of AI across the firm’s entire ecosystem and deliver greater returns as usage scales.

For knowledge leaders, the opportunity is just as significant. Institutional knowledge that has traditionally been scattered across documents and matters can become useful context for the AI tools lawyers already use — without requiring lawyers to manually assemble and upload the right files each time.

AI models and applications will continue to change. A firm’s context strategy shouldn’t have to start over every time they do.

NetDocuments provides a common layer that is governed, structured, permission-aware, and designed to work across the AI tools our customers choose.

That’s the promise of MCP with context: open choice on the AI side, trusted legal context as the foundation that powers it, and greater value from every connection.

To learn more about how this works in practice, including which tools NetDocuments supports with MCP, click here.