Pioneering law firm, Mishcon de Reya, has launched its latest venture, MDR Research, ‘a new group that brings together lawyers, engineers, researchers, entrepreneurs and technologists to advance the frontier of legal systems understanding’.
They have also launched a new funded PhD fellowship with University College London (UCL) aimed at accelerating the use of legal AI approaches. The Mishcon ‘Fellowship in AI’ has been awarded to Yao Lu, whose research will focus on NLP and adapting GPT-3 language models – potentially with a legal application. In particular it will look at how to leverage pre-existing large language models for uses they may not have been designed for, e.g. for working with legal documents and caselaw.
The firm has also joined UCL’s Centre for Artificial Intelligence’s industrial steering board, joining companies like Google DeepMind and Cisco Systems to help steward the development of AI in the UK.
In terms of the new broader MDR Research group, this is to some extent a formalisation and then extension of the work carried out at the firm already by Alistair Moore, Head of Analytics and Machine Learning, Dan Hoadley, Head of Litigation Data, and several others.
It is also a re-emphasising of the firm’s collaborative work with academic bodies. Plenty of law firms work with universities, but Mishcon wants to formalise this now into a permanent and ongoing series of projects, all housed under one roof.
Areas they want to focus on include: ‘how data is collected, interpreted and used to support the practice of law, across areas as diverse as AI and machine learning, distributed systems and graph databases’.
Moore and Hoadley told Artificial Lawyer they are working on a wide range of projects, some of which cannot be made public yet, but much of their output will be eventually shared for the benefit of the market as a whole, rather than kept just as proprietary research inside the firm.
‘For some years we have been working on quasi-academic projects and now we are bringing it all under one roof,’ they explained.
In terms of the AI fellowship, they noted that firms ‘do not want to spend $10m on a language model, so we want to use other people’s models and see if we can manipulate them’ for legal use cases.
Other projects they are working on also connect to NLP, such as one looking at caselaw analysis in relation to Judicial Review in collaboration with vLex and York University.
In addition, they have also used graph databases to better understand how courts cite other courts. Moore and Hoadley stressed that this was for more than just academic curiosity, as they were seeing how UK courts were cited by other Common Law nations, such as Singapore for example, and how those numbers were changing – giving an indication of the UK’s legal standing in the world.
Moore and Hoadley noted that this kind of research provided ‘a canary in the coal mine’ and the results may give the legal establishment a lot to consider, as the UK’s ‘influence is now waning a bit’.
They added that as more and more caselaw became easier to access digitally around the world, lawyers would shop around for whatever seemed most useful, and the automatic deference to the UK’s caselaw to provide guidance on similar Common Law matters – which had been the pattern in the past – was no longer as likely.
However, one solution to this could be, they said, to ‘develop an open-source legal infrastructure and give it away’, which could help to maintain the UK’s influence. They also noted that via the firm’s office in Hong Kong they are working with HSBC on a project in China.
This site also asked about the firm’s early work on using blockchain approaches to property transactions, to which Moore and Hoadley replied that this was ongoing work, but noted that at present the UK Government is not mandating blockchain-based conveyancing. However, they have also done work on ‘digital bills of lading and digital trade structures’, so their interest here will continue.
And that’s not even all of the projects they are working on at the moment. So, one can see why they felt the need to create a dedicated MDR Research group.
Moore and Hoadley also said that part of the MDR Research ethos was not just to link projects to academics, but to get their output published and then get feedback. This site noted that this was a more traditional scientific approach, where research and discoveries were peer reviewed and shared among a broad community of experts to validate them, with the outputs then helping the entire scientific world to move ever onwards.
And that’s one of the most positive things about this endeavour. Very often legal tech innovation – for obvious commercial reasons – is held very closely by those who have pioneered that area. However, the other side of science is not just about bagging patents, or owning products, but rather sharing knowledge for the good of all.
Overall this is great work.
—
P.S. with regard to the AI fellowship at UCL, Moore said: ‘This research is an important stepping stone in maximising the potential of technologies like AI to overcome challenges in how the legal sector currently operates. We also hope it will eventually help to remove barriers to open up legal knowledge and access to justice.’
While the PhD student, Yao Lu, said: ‘I am honoured to receive this fellowship to support my PhD research. Our initial work shows that prompt-based learning in natural language processing has great potential to solve complex tasks which usually require human and manual expertise.
‘The work reveals the order sensitivity issue of current prompt learning approaches, which is crucial for solving NLP tasks. With the support of Mishcon de Reya’s fellowship, I am confident to further explore this direction of research and produce new findings.’
And Professor Pontus Stenetorp, Deputy Director at the UCL Centre for Artificial Intelligence, concluded: ‘We are immensely grateful for Mishcon de Reya’s decision to support the UKRI Centre for Doctoral Training in Foundational AI here at UCL. Their contribution towards foundational research into AI methods is applicable both to the legal domain specifically and to the growth of the AI sector in the UK.’
You can find a link to some of the research here: https://arxiv.org/abs/2104.08786