‘Legal + GenAI’: Ample Opportunities for The Taking
How can Generative Artificial Intelligence (GenAI) help make processes smarter and faster? In almost all sectors of the economy, companies are looking for opportunities around GenAI while also being wary of the pitfalls. GenAI now has thousands of applications (‘use cases’) in a variety of fields. Besides highly visible applications such as generating fantasy images and videos like deepfakes, there are also many serious and business applications where GenAI has a role as a co-driver or assistant to a professional. This also applies to lawyers, for their offices as well as the legal processes.
Opportunities and risks
There is a field of tension between Artificial Intelligence (AI) and the legal profession. In the field of AI and data, (European) regulations are rapidly evolving. The risks of AI lie mainly in the areas of reliability (has the output been legally checked?), data protection (do AI applications handle company and personal data carefully?) and ethics (are all outcomes desirable?). At the same time, ‘legal + GenAI’ is a great combination: it is a greenfield where the opportunities are there for the taking. GenAI can both analyze and generate texts. Lawyers are concerned with texts, language and concepts unlike any other profession. At last, there is a highly accessible tool that allows one to analyze legal documents.
Lawyers not happy with legal tech
It is questionable whether lawyers are going to embrace GenAI en masse. After all, large parts of the profession are known as conservative, cautious about knowledge sharing and strongly focused on (monetizing) implicit knowledge on often complex topics. Dealing with knowledge management can sometimes encounter resistance, because knowledge sharing can impede the traditional revenue model. Knowledge management as such is not ‘billable’.
And it is remarkable that lawyers have so far succeeded in not being hindered by barely or no use of legal tech. The operation carries on as normal despite the absence of legal tech, clients do not yet walk away, and neither do fee earners see their hourly turnover threatened. Lawyers are known to be risk averse. It is important to understand that AI will not take over or define the work, but will complement, support, and accelerate it - something that is becoming increasingly notable in numerous sectors. As lawyers, we will have to use applications around specifically GenAI within a well-developed and controlled framework, to avoid creating ‘fake case law’, for example - now a notorious example in the field.
Legal tech and young generations
On the other hand, there is an influx of young people in legal practice, and especially professionals from generation Y and Z are extremely curious about this new technology. Increasingly, graduate students have legal tech and data-related subjects in their masters. They expect employers to ensure that the seniors they deal with are open to this. At big firms, tech savviness is becoming more visible.
The rise of GenAI-powered legal tech also calls for a different type of lawyer, someone less risk-averse or who dares to take position earlier, who wants to pioneer and is open to experimentation. While AI legislation at European level is gradually developing, a legal organization must now take up or adjust its position. Legislation around AI and digital requires multidisciplinary cooperation between lawyers, data-, IT- and cyber professionals.
On top of this, precisely because of the approachable nature of GenAI, clients of legal service providers are increasingly asking (justified) questions. It is widely known and almost everyone has seen examples of the output of GenAI. How can you get an answer to a complex question faster? How can you deploy teams of juniors in a smarter way? With the advent of GenAI, lawyers will have a tool in their hands that allows them to have not just a subset of case law, but all case law co-researched at negligible cost. Previously, that was not doable. That ability is rapidly becoming more important as legislation becomes increasingly inter jurisdictional - relying on what you have read or heard yourself is becoming less and less effective. Here too, GenAI can contribute to the necessary enrichment and deepening of the practice of lawyers and jurists.
For the internal legal function of companies - often a cost center - GenAI also offers opportunities. It can reduce dependence on external lawyers, simply through increasing in-house capacity and capabilities.
How do you deploy GenAI?
Now, the big question is how to successfully deploy GenAI in legal practice. Despite that low-threshold nature, it is not the case that, as a lawyer, taking out a paid subscription to a GenAI application is enough. On the contrary: it will require stringent measures to deploy the technology properly and securely. The good news is that legal Large Language Models (LLMs) are already available within the hundreds of specialized LLMs on which GenAI works. This makes it possible to execute various tasks and processes faster or more completely and smarter. Think of large-scale analysis of contracts, which is relevant in large and complex transactions such as M&As. Here, GenAI can play a useful role in finding liabilities in due diligence processes.
Another example with which quite a few successful pilots have already been carried out, is looking for nuances in case law: this can contribute to better client advice. A third example is ensuring compliance in opinions when it comes to European harmonization. And finally, significant strides are being made in solutions that can produce advance estimates on the feasibility of procedures, based on different choices and arguments.
New roles and competences
AI offers new opportunities and risks in the non-legal business processes, requiring legal professionals to adapt and develop new skills in using AI. The use of GenAI requires new capabilities within organizations. Lawyers, like any other professionals, will have to learn to work with GenAI. This includes controlling algorithms (‘prompt engineering’) and interpreting and assessing AI output. Those who do not know what prompt engineering entails will continue to ask stupid questions to a smart system. It is also important to be able to recognize manifestly incorrect output. In addition, the work processes of the legal function are changing. IT is going to play an essential role in both the primary process and data protection.
To work more data-driven has implications for data management: the basics, such as the data quality, IT security and compliance, need to be in order. This also calls for more T-shaped and O-shaped professionals (respectively, professionals who combine in-depth knowledge in one area with good basic knowledge in multiple areas, and professionals who combine their professional knowledge with knowledge of people, clients and business, ed.) who can work together effectively in multidisciplinary teams. For the lawyer, this means that he or she will have to appropriate the IT and data domain. Those who succeed are also more likely to be successful in using legal tech in their own practice. But honesty dictates us to concede that we - as O-shaped legal professionals - are often not there yet.
Take action now
Anyone who does not delve into the fields of AI and GenAI will find it quite challenging to catch up in a few years. It is crucial for legal professionals and organizations to act now and integrate AI into their business activities and processes. Where to start? Make sure to upgrade your models, your data, your systems, and your people. But also avoid turning lawyers into programmers. Ensure that the cooperation between disciplines in the organization is strengthened. Embrace and explore the technology so that you can gain insights into opportunities and pitfalls in a timely manner. Adopt an agile approach: take small steps so you can adjust timely and stay well on track.
Essay by Dominique Poot, Director Legal Management Consulting at Deloitte Legal, and Sebastiaan ter Wee, Partner Tech Law at Deloitte Legal. Published in Management Scope 01 2024.