Part 1 - AI for legal professionals: Where to start?
The more lawyers use artificial intelligence (“AI”) to deliver legal services, the greater the need for them to understand what underpins the technology. The first instalment in our series on AI for legal professionals, this article is the first in a series designed to support lawyers in understanding AI and using it responsibly. We begin by answering the fundamental question of what AI actually is.
Where to start?
In public discourse, “AI” is often used loosely, describing anything from chatbots to ChatGPT, which leads to confusion what AI actually is and does. As AI tools become increasingly present in legal practice, it is important to be able to distinguish between the various terms used and the technologies underlying AI.
AI broadly refers to algorithm-based technologies that can perform tasks which typically require (or previously required) human intelligence. These tasks include pattern recognition, language comprehension and problem solving. Although AI often evokes futuristic images of autonomous machines, most of the AI systems in use today fall under the category of “narrow AI” (tools designed for specific, well-defined tasks). These systems power everything from predictive text in email applications to the software that scans legal documents for relevant clauses. In contrast, "general AI", which would exhibit the full range of human cognitive abilities, remains a theoretical concept and is not yet realised in practice.
Various algorithm-based technologies are included under the narrow AI umbrella. At the core of many modern AI systems is “machine learning”, a subset of AI that enables systems to learn patterns from data rather than relying on hardcoded rules. Within machine learning, “deep learning” refers to the use of layered neural networks that excel at handling complex data like images, audio and large volumes of text. A key application of deep learning is “natural language processing”, which allows machines to understand, interpret and generate human language.
Finally, “generative AI”, a form of AI which creates new content such as text, images and code, has gained recent prominence. Generative AI tools, such as ChatGPT, rely on “large language models” (“LLMs”), which are deep learning models trained on vast amounts of text to predict and generate language in a coherent, contextually appropriate way. While the capabilities of LLMs can appear human-like, they do not “understand” content in a cognitive sense which is why it is important for lawyers to grasp the underlying mechanisms and limitations of these technologies.
Together, these technologies support the development of systems that can review and summarise documents, flag inconsistencies and missing clauses, and even conduct basic legal research. More information on AI’s application in dispute resolution in particular can be found in our Lexis Nexis article on the topic.
Comment
While AI can accelerate research and streamline routine legal tasks, it is not a replacement for legal judgment or professional responsibility. Lawyers must therefore maintain a clear understanding of how the technology works and its limitations.
As this series continues, we will explore the use of AI in various areas of legal practice, from e-disclosure to case analytics and prediction, with a view to supporting lawyers to understand AI and use it responsibly. In the next edition, we will delve into the fascinating topic of AI hallucinations and their implications on legal practice.
About Hausfeld’s Digital Markets expertise
The Hausfeld Digital Markets Hub