Our Take
MCP is no longer a technical standard — it's a procurement decision that will sort firms that deploy usable AI from those stuck with capable-but-isolated tools.
Why it matters
Most law firms waste the productivity gains from AI by using lawyers as manual bridges between disconnected systems. As iManage, NetDocuments, Harvey, and Legora all shift to MCP support, the next 18 months will determine which vendors firms lock into.
Do this week
General Counsel or Head of Legal Tech: audit your current DMS, transaction management, and matter management vendors for MCP server roadmaps before the next procurement cycle so you can avoid locking in systems that won't interoperate with AI.
Two systems integration problems are slowing AI adoption
Most law firms now run at least one generative AI tool in production. But underneath the announcements from Harvey, Legora, iManage, and NetDocuments, a structural constraint is blocking real productivity: the systems don't talk to each other.
The first problem is the context gap. An AI tool can summarise a document or draft a clause, but it cannot always see the rest of the matter — the precedents, the related correspondence, the deal-specific instructions elsewhere in the stack. Lawyers manually pull that context together before every AI interaction, consuming much of the time the AI saves.
The second problem is the action gap. Even when the AI produces something useful, it usually cannot act on it in the systems where work actually happens. It can draft a status update but not post it to the deal room. It can flag a missing condition precedent but not update the closing checklist. The lawyer becomes the messenger between the AI and every other platform.
These two gaps explain why many AI pilots plateau at "useful, but not transformative." The model is not the bottleneck. The connectivity is.
MCP is now a strategic procurement decision, not plumbing
The Model Context Protocol, an open standard originated by Anthropic and now backed by expanding vendor support, is designed to close both gaps. Think of it as HTTP for AI-to-system integration. Instead of each AI tool requiring custom integrations with every legal system, all vendors connect to a common interface.
iManage launched its MCP server in May. NetDocuments is following. Harvey and Legora are positioning around agentic workflows that depend on exactly this kind of connectivity. Vendors that are not there yet will soon face hard procurement questions from large customers.
Five use cases are credible now: pulling full matter context into AI sessions so drafting is grounded in the actual deal; letting AI assistants see live status across transactions and write back updates; connecting AI directly to data rooms so issues flow into reports; making firm precedents available to the AI in structured form; and pulling status, financials, and risk into client-facing updates without manual compilation.
None of these work at scale through copy and paste. All depend on the AI being able to read from and write to the systems where work actually happens.
The question in front of firms is not whether to take MCP seriously. It's whether to do so now, while the market is still forming, or later, after others have set the pace. Systems bought, renewed, or decommissioned over the next 18 months will either be MCP-enabled or they won't. The AI tools deployed will either be able to act across the stack or they won't.
Audit vendor roadmaps before the next contract renewal
Procurement teams should ask every system vendor — DMS, transaction management, matter management — whether MCP server support is on the roadmap and when. Get a written commitment, not a verbal one.
For the AI tools your firm deploys, confirm MCP client capability. This is now table stakes for any tool claiming to handle real workflow.
Identify one high-value integration to pilot. The five patterns above (matter context, transaction coordination, due diligence, precedent, client reporting) are all viable starting points. Pick the one that saves the most lawyer time or unblocks the fastest deal cycle.
Brief partners on what changes when the AI can finally do the work, not just describe it. The productivity story shifts from "this saves draft time" to "this eliminates the manual handoff entirely." That is the conversation that changes deal profitability and staffing.