Shining a light: Using AI to transform the Legal Sector
Read what the members of our AI for Services network have to say about the challenges and opportunities in the Legal Sector.
Supported by KTN, AI for Services is a UK wide network bringing together AI & Data high growth entrepreneurs with leading professionals and academics working in the high professional services sectors.
This month we interviewed three of our member organisations that were successful in securing an Innovate UK grant through the Industrial Strategy Next Generation Services Challenge Programme: Legal Beagles, Solomonic and transparently. We asked them about their projects, their motivations and what the future of the sector holds.
Legal Beagles: Global Access to Justice via AI & Community
This project seeks to provide easy access to legal answers, support and community advice enhanced by AI technology, which can analyse and scale knowledge to predict best routes for consumers to find solutions to legal issues. The objectives are to provide quality legal support in a cost effective manner, to a global population that is increasingly struggling to gain access to justice and consequently researching online to find answers to legal matters.
We interviewed Pamela Austen, COO of Legal Beagles
Why was your project developed? What was the inspiration and aim?
In a nutshell we are part of the solution to provide access to justice for individuals who need support identifying and dealing with legal matters. Our project was developed to take the power of AI and a community based forum, to machine learn frequently asked questions, specifically in our employment law area as the Proof of Concept. The end result is to have a chat bot able to identify, and either answer the questions or provide relevant signposting to guides and templates and then if necessary pass to subject matter experts for further advice.
What have been the challenges so far? How did you overcome them?
The main challenge is that our technology partner (IBM) recently pulled out of the project and we have had to find a new technology provider. Our Monitoring Officer and Innovation Lead have provided us with great support and we are in the final stage of being given the go ahead to restart the project with a new provider, with whom we are really confident we can successfully complete the project.
What does the future look like?
The legal market is a buoyant market and will continue to be so, however in terms of technology developments these are primarily aimed at improving processes and procedures for corporate law firms and in-house counsels. There is very little that is benefiting individuals in the every-day legal situations they are facing. We believe that the way in which legal services are provided to this sector of society needs to and will change over the next few years.
Solomonic & Warwick Business School: A litigation machine learning project
Solomonic is the UK’s leading litigation analytics company. They produce rigorous and valuable analysis of high court decisions to inform high value litigation strategy and decision-making. The cost of doing the analysis of decisions is significant and relies on a small set of qualified experts.
We interviewed Tom Jewkes, COO of Solomonic
Why was your project developed? What was the inspiration and aim?
The project was established to see just how much of the analysis of complex high court judgements we could get a machine to do and do more accurately than a human to make us more efficient and radically expand our capability.
Our project had two aims, to improve our efficiency and to enhance our capabilities and our global competitiveness. Solomonic draws on a small talent pool of qualified solicitors and barristers for the complex and accurate analysis our methodology demands. This makes the production of quality data slow and our capability to expand our data set is limited and consequently our competitiveness. At the same time Warwick Business School Business Systems department had a deep interest in developing a sophisticated NLP approach to analysing complex unstructured texts and seeking to answer questions of those texts requiring interpretation or inference.
What have been the challenges so far? How did you overcome them?
Probably the biggest challenge was data. To train a machine in the basics you need to give it
accurate data and in sufficient quantities. Luckily, we had done a lot of that work already, but it was an ongoing struggle to keep feeding the team (and the machine) with enough of the right data to progress our development. More generally we had to work in partnership which required us to keep focused and maintain very regular and, especially in the early part of the project, quite structured communications.
What does the future look like?
Covid-19 will impact the legal sector in the short term in a negative way, but in the medium to long term, major economic and social disruption tends to generate legal disputes and we anticipate this happening in the near future. We plan to use the tools we have developed in this project to help us provide high value insight and intelligence to the law firms and the organisations who get into these disputes and improve the outcomes for them.
Transparently: Placing integrity at the heart of all negotiation
Transparently’s project wished to gain an understanding of the complexities and technical issues associated with application of AI techniques to the analysis of conversational data. In the longer term the company is seeking to introduce a range of software tools, to augment the role of legal professionals during any discussion, negotiation or dispute resolution.
We interviewed Steve Stewart, COO and Becki Cassia, CEO of transparently
Why was your project developed? What was the inspiration and aim?
Our focus at the moment is family law; specifically, we are looking to support all processes associated with separation and divorce. As a company, we are passionate about promoting integrity and reducing conflict wherever possible; particularly to reduce acrimony during proceedings, to protect vulnerable individuals and to minimise the need for matters to be resolved in court.
In support of new legislation, our interest was to investigate how coercive and controlling behaviours may be detected, so that we could gain the know-how necessary to support next generation services. Our intention was to reduce natural bias and to identify deliberately manipulative traits. Ultimately we wish to provide a machine-supported ‘second opinion’ to legal professionals during emotional and difficult negotiations.
What have been the challenges so far? How did you overcome them?
We designed our project to address four fundamental challenges:
- the identification of the most appropriate subject matter experts, academic institutions and partners; balancing AI / Machine Learning (ML) and intent analysis experience
- to understand how AI and ML techniques could be applied to free-flowing conversational data; complex, variable context, random and unstructured
- to gain knowledge of the effect that operating within the highly emotional context of separation and divorce, could have on such techniques and any bias that may be introduced as a consequence
- to begin to understand the structure of the platform that would be required to identify behavioural traits, identify them linguistically and then analyse the resulting data securely and ethically.
Our project concluded successfully; using a small, real-world data set, we were able to demonstrate that such linking and analysis is possible. The next step for us is to develop the research further, given sufficient resources, towards an initial prototype and then on to developing a full commercial solution.
What does the future look like?
In no way to minimise the personal tragedies experienced by many during the Covid-19 pandemic, the crisis has impacted the legal sector profoundly, in ways that will be seen both positively and negatively.
In the short term, the disruption to the Courts, to office-based law firms – traditionally wary of cloud based technology and built around in-person meetings, paper systems and wet signatures – there will be significant disruption to both legal professionals and private clients.
In the longer term, exposure to technology as required by the current restrictions and the streamlining of systems to focus on what truly makes the difference, will transform an entire profession. We believe that adoption of new technology, which previously would have taken a decade, will be greatly accelerated.
Integrated end-to-end solutions, that fully address the issues exacerbated by the Covid-19 crisis, which protect vulnerable individuals, improve access to justice and reduce cost, will be best placed to succeed and be adopted by the profession. The trick will be to keep the wheels turning until then.