This blog post highlights some key points from this workshop.
What are the opportunities?
- Access to justice: Machine learning could be used to provide legal advice to individuals or small businesses who cannot currently access legal services. The DoNotPay chatbot that I’ve blogged about previously is one example of how machine learning can guide individuals through legal processes. Another is Amiqus, a start-up that builds tools to give individuals open and free access to legal information, with the objective of enabling them to make informed legal decisions.
- Corporate efficiency: Machine learning could help save time and money by streamlining business processes. For example, machine learning could carry out early due diligence work, by scanning the full range of due diligence papers rather than a 10% sample as lawyers would. The application Kira uses machine learning for contract analysis to uncover relevant information.
- New insights: Machine learning could provide ‘added value’ by creating new insights from legal data and facilitating better interaction with clients through new services. For example, Cognitiv+ has developed a machine learning platform that provides legal and regulatory insights to law firms and businesses, enabling them to better understand and act upon their compliance obligations.
What are the barriers?
Some of the barriers to using machine learning in the legal sector echo those noted in our work with other industry sectors. These include:
Access to data: Data is seen as the intellectual capital powering bespoke legal services. The law firm model therefore often discourages information sharing based on intellectual property concerns. In addition, clients often do not want their data to be shared beyond those working on their case. Client consent could therefore limit the use of machine learning, even in legacy cases, as law firms might lose the data after client contracts end.
Business models: The use of machine learning to automate time-consuming tasks such as basic research could drive changes in the legal sector’s traditional hourly fee-based business model. Similarly, for those who already have access to legal services, human interaction is often preferred as a fundamental part of the service, even if it is more expensive or potentially less rigorous. The desire for such interaction could be a key influence on how machine learning is used.
Skills: As in many sectors, machine learning is creating new skills needs. Where technology was previously a peripheral activity for lawyers, machine learning could sit at the centre of their activities in future. Professional legal culture is steeped in history and may underpin some reluctance to engage with technological approaches to the law, even at university level.
Machine learning, society and the law
A number of machine learning applications are becoming increasingly familiar to lawyers; these typically use machine learning to process large amounts of data from previous cases and detect patterns in this data. As these applications gain prominence, it will be important that they are used in a way that maintains space for creativity in interpretation of the law, rather than solely repeating precedent, so that the legal system remains able to evolve.
For a service sector such as law, client demand might end up being the key factor which drives the uptake of machine learning. Some clients may expect to see this, or similar, technologies used as a matter of course, and even request their use as a matter of business efficiency. For others, or for certain circumstances, human interaction may remain central to what law firms offer.
As the role machine learning applications play in day-to-day life increases, demand from clients may also change. As a result, the role of the lawyer will evolve, with increasing emphasis on strategic thinking and bespoke client services, in addition to new technical skills. This makes an exploration of the potential impact of machine learning in the legal sector both timely and important.
Our interest in machine learning does not stop here, and we will be continuing to explore visions of machine learning in the UK as part of our wider project on the technology. For further information, take a look at our website.