How to adapt a law firm for AI

First of all, it starts with a conviction that AI is going to fundamentally change the legal profession and their business but realising that it’s too difficult to say what exactly the nature and shape of this will be.

What we always say to firms is take it in three stages. Stage one is to experiment and then, based on that experimentation, think about how to optimise what you’re doing today. Stage two is looking at how to engineer these AI solutions into your legal workflows and what that means for pricing, billing and staffing models. Then stage three is thinking about changing your business model and how you should redesign your service portfolio to make sure it fits with where the value is.

A lot of folks are thinking about whether to buy or build. That’s the biggest strategic choice you need to make. My conclusion is that it is pretty much impossible for a law firm to organically build the best AI capabilities. There’s too much innovation in the market, and most law firms will be unable to keep pace or hire the best tech resources. So there’s a big question mark around the quality and the long-term viability of whatever you build yourself.

The best strategy is to partner with a vendor who’s really investing in and has a long-term view on the market. Another consideration when looking for a vendor is to be sure that they are offering professional-grade solutions, and that they can stand behind the security and ethical considerations, as well as performance and scalability.

This is really important, because a lot of law firms are rolling out AI with good intentions, but the adoption at an individual lawyer level sometimes doesn’t match that intention. So first and foremost, it’s about explaining to lawyers why it is important for them to be trained on AI and what is at stake for them. But equally, it’s also about giving them some practical assistance and guidance to bring those people on the journey. A lot of the time, it is about seeing, trying and using AI – and once they’ve done that, they’re hooked. So it is that initial barrier that people struggle with.

You need someone who can create a technology blueprint for what the law firm is rolling out. That’s very important, because there are hundreds of AI tools out there. If you’re not careful, very soon there will be a hundred different tools across the firm, and lawyers need to do one task in one tool, one task in another tool, and so on.

That’s why you need somebody with a technology blueprint, which outlines how the organisation is simplifying everything and making sure it’s investing in the right tools. However, if you only rely on these central roles to be successful with AI, it’s not going to work. AI is about democratisation. So ultimately, every practice leader and every partner needs to be a role model and show how they are using technology and how it is helping them – that’s how it propagates.

There’s a generational aspect to take into consideration – if you’re a managing partner or at a very late stage in your career, it’d be ideal for you to be hands-on with the technology, but it’s okay if you’re not. But you do need to understand the business value of it and to be able to articulate it in a very crisp and clear fashion. If the leaders are not bought in, people can sense it.

For new associates or senior leaders within practice areas, there is no choice – you have to be hands-on with the technology and embrace it. But while there are perceptions that older lawyers won’t use AI solutions, our research shows this is inaccurate – we are finding there is no statistical correlation in terms of age and the use of AI tools.

This is a really important aspect because if everybody has the same standard AI technology, then how do you differentiate? From a technology perspective, the differentiation comes from the fact that you have this treasure trove of data from all your previous work. Therefore, you need to mine that data to feed the AI engine and put your own flavour on the answers AI produces.

A lot of firms say they have amazing data repositories, databases and document stores, but if you ask if they can find the right documents to connect to the AI at the right time, the answer is often no. That’s why having the right data foundations is an important part of an AI strategy.