What MCP Is and Why Legal AI Needs a Standard
The Model Context Protocol (MCP) is an open interoperability standard that lets AI applications consistently access data and actions from other systems, so legal AI tools can read and update work where it lives instead of relying on manual copy and paste. Most law firms now run multiple generative AI tools in production, but those tools typically sit apart from document, matter and transaction systems. That separation creates a context gap, where AI cannot see the full picture of a deal, and an action gap, where it cannot update the platforms lawyers use every day. MCP addresses both gaps by defining two roles: MCP servers expose systems like a DMS or transaction platform through a common interface, and MCP clients are AI tools that consume those capabilities. One MCP-enabled stack can then connect to many AI law firm tools without custom integration projects.
Closing the Context and Action Gaps in Law Firm Workflows
Inside most firms, documents sit in a DMS, matters in a separate management platform, and transactions in dedicated deal systems, while AI tools operate in yet another interface. Lawyers become the manual bridge, fetching precedents, correspondence and deal instructions so AI can draft or summarise, then retyping AI outputs back into checklists, deal rooms and client updates. This is why many pilots stall at “useful, but not transformative”: the model is capable, but connectivity limits impact. MCP-based integrations aim to close this by letting AI assistants pull complete matter context and then act on outcomes. With MCP, an AI could see a live closing checklist, flag missing conditions, and write updates directly into the transaction workspace. Document and matter context, due diligence flows, knowledge access and client reporting all benefit when AI tools can read from and write to the systems where legal work is actually done.
Frontier Legal AI Vendors Are Standardising on MCP
As MCP gains ground, leading legal tech vendors are starting to treat it as the default way to integrate. Harvey is expanding workflow agents that depend on direct access to live deal data, while Legora has announced an Agentic OS that assumes deep, protocol-based connectivity across the legal stack. Document management providers are also moving: iManage launched MCP support on 14 May and NetDocuments is headed the same way, turning these platforms into MCP servers that any compatible AI client can use. According to Artificial Lawyer, this shift makes MCP “a procurement question” for firms evaluating new systems over the next 18 months. Tools that can act across an MCP-enabled stack will be easier to justify, because they avoid brittle, one-off integrations and unlock more complete MCP legal AI integration scenarios, from automated matter summaries to cross-platform transaction management.
doola and Vercel: Legal Infrastructure Embedded in Developer Tools
MCP’s impact is not limited to law firms; legal infrastructure automation is starting to appear inside mainstream developer platforms. doola, an AI Business-in-a-Box provider, has released an MCP integration inside Vercel, making it the only company formation platform embedded natively across Claude, Replit, ChatGPT, Lovable, Perplexity and Vercel. Through Vercel’s AI-native interface v0, founders can form a U.S. LLC by conversation without leaving their deployment environment, while doola handles filings and routes them to a dashboard for tasks like EIN setup, banking and compliance. Arjun Mahadevan notes that “with doola MCP inside Vercel, you can push code and form the company in the same workflow.” This is a concrete example of MCP turning legal actions—like incorporation—into steps inside the same AI workflows developers already use, rather than separate, manual processes.
What MCP Adoption Means for Legal Operations and Business Teams
As MCP spreads, it is reshaping expectations around legal tech interoperability standard choices. For law firms, the decision is whether to treat MCP as background plumbing or as a strategic layer that guides procurement. Systems purchased or renewed soon will either be MCP-enabled or isolated, and that choice will decide how easily new AI law firm tools can plug into document, matter and transaction platforms. For in-house teams and startups, MCP-based tools like doola on Vercel show how legal steps can become part of product and deployment workflows instead of blocking them. The more MCP servers and clients enter the ecosystem, the less friction there will be between legal operations and core business activities: AI can fetch the right context, take actions in the right system, and keep legal workflows in step with how products are built, shipped and run.
