What Is MCP and Why You Need It – Artificial Lawyer

By Antti Innanen + Elias Ylönen.

Have you ever been frustrated with phone chargers? You need a Lightning cable for your iPhone, but don’t have one at hand. Then you land in London and discover you need yet another adapter. Damn.

Wouldn’t it be easier if there were just one universal charger?

That is the situation AI agents face today. There is no shared language. Each system is isolated, and every new data source requires its own custom integration.

But this is starting to change. The ‘USB-C for AI‘ has arrived. Just as USB-C provides a single standard for connecting devices to chargers and accessories, the Model Context Protocol (MCP) provides a standardized way to connect AI models to data sources and tools.

Introduced by the AI company Anthropic, MCP is already becoming the default standard for AI interoperability. Microsoft, Google DeepMind, and OpenAI have all announced support. In April 2025, Demis Hassabis, CEO of Google DeepMind, described MCP as ‘rapidly becoming an open standard for the AI agentic era‘.

For legal tech vendors and law firms, the potential is wide open. MCP makes it possible to connect systems that were never designed to work together, unlock new workflows that save time and reduce friction, and create entirely new categories of legal applications.

And yes, I wish I had come up with the USB-C analogy myself. Credit goes to Anthropic. The adapter struggle is real though, especially in London.

The End of Integration Moats

For decades, legal tech platforms built competitive advantages through proprietary integrations. Document management, practice management, and billing systems became walled gardens where complexity itself was a moat.

MCP blows up that model. Any AI agent can now connect to any MCP-enabled platform through a standardized protocol.

iManage’s early adoption in August 2025 shows they see where this is going. As CEO Neil Araujo explained, the goal is to let AI ‘automatically discover available content and capabilities in iManage Cloud, while respecting existing user permissions and access controls’.

Just last week, Juro and Wordsmith announced a partnership (see AL article here) using MCP to provide an integrated AI offering to their clients.

These early moves show what is possible. As more vendors and law firms embrace MCP, new partnerships, integrations, and services will follow. Competitive advantage will increasingly come from being open, connected, and ready to collaborate.

So, How Do You Build These Things?

The easiest way to get started with MCP is to enable tools in an LLM like Claude. Right away you can see how tool use helps AI fetch new data and become more useful.

To go a little deeper, think of it from two perspectives. If you want to expose your data to AI, you build an MCP server. If you want to let AI access data, you build an MCP client.

We build these kinds of integrations for enterprise customers, helping them connect their systems and unlock new workflows. But one project we built for ourselves illustrates the power of MCP especially well.

We connected Outlook Calendar and Apple Calendar to Claude. With that integration, you can ask Claude to ‘recap my week‘ or ‘find the best events in Berlin this weekend that fit my calendar’.

Calendar integration is one of those practical use cases that genuinely improves daily workflows. We made it open source, so you can try it out yourself.

This shows exactly what MCP makes possible. Without the connection, Claude has no idea what is happening in your calendar. Once you link them, it does. And if you take it further, you can connect AI tools to almost any data source.

What Are the Risks?

There are some real dangers to consider. MCP connections should be treated like any other system integration, with proper security measures, limited permissions, and careful thought about what data you expose. And even without malicious intent, things can go wrong.

Because MCP can also allow AI to take actions, not just read data, there is a risk of accidental changes. If connected to a database, an AI system might delete records, modify files, or send emails without approval. Even with the best intentions, it could misinterpret instructions.

It is important to stress that this action-taking capability is optional. MCP does not mean handing over full control. It simply provides the option for systems to act on your behalf if you choose to enable it. Still, this makes MCP higher stakes than simply letting AI read documents. Once an AI can act inside your systems, the potential impact of mistakes becomes much greater.

And, like any integration, you are also depending on external services to remain online and stable. These are all risks you can mitigate, but do not just plug anything into your AI tools.

The Opportunity Ahead

MCP is not about replacing platforms or discarding what has been built. It is about making everything work together more easily. That shift creates space for law firms and vendors to reimagine what their tools can do, who they can partner with, and how they can deliver value.

Who could have predicted Juro and Wordsmith suddenly finding themselves in a Taylor Swift and Travis Kelce–style romance? The opportunities for collaboration are just as unexpected, and they could grow into something big.

The field is wide open. Now is the time to explore it, to partner, and to build.

—

[ This is an educational think piece for AL. ]

About the authors

Antti Innanen is a tech lawyer, legal design enthusiast, and an AI geek. His new book “Prompted: How to Create and Communicate with AI” will be published by Routledge in 2025. You can find more about Antti and his work here: LEGIT, Dot, and Legal Design School.

Elias Ylönen is a software developer and AI specialist. He is the CTO at LEGIT, where he focuses on building the agentic future. Elias contributed the technical insights for this article and has already earned a reputation among the team as an MCP wizard.

—

Legal Innovators Conferences in London and New York – November ’25

If you’d like to stay ahead of the legal AI curve then come along to Legal Innovators New York, Nov 19 + 20 and also, Legal Innovators UK – Nov 4 + 5 + 6, where the brightest minds will be sharing their insights on where we are now and where we are heading. 

Legal Innovators UK arrives first, with: Law Firm Day on Nov 4th, then Inhouse Day, on the 5th, and then our new Litigation Day on the 6th. 

Both events, as always, are organised by the awesome Cosmonauts team! 

Please get in contact with them if you’d like to take part.