Professionals will train for many years to become proficient and recognised in their respective, specialist fields, such as navigating the complex parameters of the law and practicing its application across a myriad of potential situations. It is these specialist skills for which we pay a premium. These elements have for some time been deemed to be so complex that automation has been kept at arm’s length and as far in the future of the possible, due mainly to the complexity and systems set out so far to tackle the challenges.
We have seen a lot of AI tools appearing in the Legal sector, but so far, and perhaps as a symptom of what is considered possible, it has only been within the spectrum of machine learning and pattern matching to identify sets of anomalies or trends. These are then used to feed in as triggers for a human to interrogate and explore further for considered resolution.
This spectrum of complex and unique interactions through to simple and automated processes reflects what we have seen in the use cases for AI across all sectors. Other industries, especially financial services and the automotive industry, are showing us shifts in adoption and are demonstrating what is possible with AI.
It is time to explore AI’s value for law across this spectrum.
Products designed for contract digestion and entity matching are already in the marketplace and some have been structured, in a number of cases, into specific products for this very purpose. The value here is two fold:
These are both examples of where a human would traditionally be asked to read and identify elements and highlight sections for wider review; not a specialist or trained skill, but instead a more editorial and generalist activity that is currently being performed by highly trained individuals whose time would be better spent on tasks aligned to their own knowledge expansion and the training of an automated system.
As we progress through the spectrum we move towards common queries, including the more frequent but less complicated elements of doing business as a law firm. For example, answering questions such as “Do I need an NDA?” are already being tackled and explored with AI by legal firms to support the areas of their clients needs where there is no longer a traditional fee to be earned or where the shift in specialism means the engagement is now less appropriate to the core skills of the team.
So here is the challenge, if we are going to leverage more of the specialist skills, we need to take a look at the categorisation of what is deemed to be simple and complex in the legal environment with a fresh cognisance of the technical possibilities. There needs to be a recognition that law firms must further embrace AI, to support in the leveraging of their skills and business in order to best serve their clients, as the environment that they’re working in puts further pressures on them operating within a wide remit.
AI already exists to answer simple questions such as the example above regarding whether an NDA is required but that is not a question that a lawyer would set out to answer, or in fact would be asked. However, we need to acknowledge that a lawyer has two core elements to what they provide when delivering a service. Firstly they are the bastions of a legal understanding but secondly, and perhaps more importantly, they are supportive advisors to the people they serve. The question a good professional would ask when faced with the example of an NDA is ‘What are you trying to protect” and “why is it important to you” rather than informing them of the process to protect themselves, and for that there is a need to move away from patterns and data and switch to codifying of the human approach and decision making process.
If we codified the human approach for this scenario then we would need to ensure that the response is 100% accurate so that it provides the same answer that a lawyer would have devised. The machine itself will assess and diagnose, as well as decide the appropriate steps and process for the client to take once it has identified their intent and salience towards each answer given.
The machine itself will assess and diagnose, as well as decide the appropriate steps and process for the client to take
These ‘inference’ or ‘decision’ platforms are already being used in regulated environments and the legal profession. The attractive element for these industries is that they not only have the ability to reach a decision and provide an answer, but can also replay this interaction back to a user or an auditor to demonstrate how and why they provided that response.
This is why a progression across this spectrum can, has and will continue into the increasingly more complex business scenarios and will allow for the true talent of legal professionals to be seen as it supports an automation shift of tasks which are deemed to be transactional away from those which are deemed complex.
In spite of the benefits associated with the implementation of AI in modern law firms, there remains a reluctance to adopt technology. Typically, lawyers and law firms are naturally risk averse. There’s a pressure to develop and demonstrate trust in everything they do, and the problem is that a technology they can’t see is often considered a big ‘no’ in terms of trust.
This lack of trust is putting the sector at risk of becoming behind the times. It is very people-centric, and so firms aren’t naturally inclined to implement much technology. The rise of cognitive reasoning solutions for law firms enables visibility of the interactions and the decision making process, and should enable trust in technology.
Therefore, the challenge to implementing AI in law firms is not that the technology needs to evolve, but rather that it’s the thinking and vision of the people who hold the knowledge which requires progression. Rather than fearing a reduction or limitation to their standing, they can express a wider organisational value and likewise increase their personal value, through the use of AI tools.