In this episode of ITWeb TV, Nerushka Bowan, a tech law expert and founder of the LITT Institute, discusses how LLMs and generative AI tools such as ChatGPT have become a double-edged sword for lawyers, as well as how it could affect jobs in the legal profession. She also provides a sneak preview of her presentation at the upcoming ITWeb Security Summit. #ChatGPT #GenAI #TechLaw #QuantumComputing #CyberSecurity
Generative artificial intelligence (GenAI) tools like ChatGPT are shaking up the legal industry, transforming everything from contract drafting, to legal research and challenging long-held notions of how legal work gets done.
This is the word from Nerushka Bowan, a tech law expert and founder of the LITT Institute, in an interview with ITWeb TV.
Bowan was speaking ahead of her presentation at the upcoming ITWeb Security Summit, where she will talk about how quantum computing will impact the world of cyber security.
In the interview, Bowan explores the promising and problematic ways lawyers are using GenAI tools.
Before the LITT Institute, Bowan worked at law firms such as Norton Rose Fulbright, where she was head of technology innovation.
“It has been such an interesting development since we got ChatGPT a couple of years ago. It has been interesting to see how lawyers have embraced GenAI. We are talking about a fraternity that has historically been technology-averse. It has been a long road to get people to use technology,” she said.
“However, what we have seen with GenAI is people are using it even when they are not supposed to be using it. That’s how much they actually love it. What has happened is because they don’t understand how it works and what the risks are, they are finding themselves in hot water.
“We’ve had lawyers from all over the world that have jumped on the bandwagon; they’ve heard about this amazing tool called ChatGPT and they’ve turned to it for things like legal research, or to find cases. They ended up citing those cases in court documents. When they were asked to produce those documents in court, that’s when they realised that the cases did not exist.”
Bowan explained this is because these were hallucinated cases, which were completely made up, and the lawyers did not understand that ChatGPT could do something like that.
She pointed out that on one hand, there has been strong interest and encouraging adoption of GenAI within the legal community. However, a knowledge gap remains around understanding its appropriate applications – what it can and can’t be used for.
“On the other hand, we have what I can call vertical AIs that have been created – tools built on top of popular GenAI tools that are focused on legal use cases, like contract drafting, due diligence and contract reviews. We are seeing a lot of law firms licensing these tools and using them in their working environments.”
She noted that while a handful of forward-thinking lawyers and law firms are already reaping the benefits of GenAI, widespread adoption across the legal industry remains limited.
According to Bowan, GenAI has created an innovation conundrum for the legal fraternity, which has historically sold time by billing per hour. For example, if a contract takes 10 hours to draft, they would charge that per hourly rate.
“Now, if I am giving you a tool that is going to enable you to draft the same document in an hour – from a lawyer’s perspective, the question is ‘how am I now going to be able to charge a client?’ So, there is more than adoption that needs to happen in the legal fraternity. We also have to relook things like the business model – how we charge clients as we adopt these tools.”
She urged lawyers to familiarise and empower themselves with how GenAI tools work first, before using them for client work.
“If you understand how a tool works and what its limitations are, then you will be able to use it in a much safer way, where you know that hallucination is an issue. You will not blindly rely on the information that it gives you because you will fact-check it before you use it in client work. These are large language models that tell you what you want to hear, but they are not necessarily trained for high levels of accuracy.”
Nerushka Bowan, tech law expert and founder of LITT Institute. (Photograph by Lesley Moyo)
She also highlighted that just like in other industries, there are fears of AI-induced job losses in the legal fraternity.
“The fears are about how will junior lawyers be trained if the AI is drafting everything? That fear is not misplaced, but I do think, as with any technology or any digital transformation that we go through, we will not completely replace humans. Sometimes we need the comfort of a human and getting a human to solve your problem – that’s what lawyers are there for.
“We are there to provide empathy, to connect with the individual, and to build relationships over a long time.
According to Bowan, in a growing number of cases around the world, individuals are turning to ChatGPT and other GenAI tools to assist in representing themselves in court.
With limited access to legal counsel due to cost or availability, self-represented litigants are using AI to draft legal arguments, understand court procedures and even prepare cross-examination questions.
“There are some tools already in the US. For example, if you have a parking ticket, you will not be able to afford to hire a lawyer to defend you. There is a tool called DoNotPay, where you can chat to that tool and it can draft court papers based on the information you provided it with and you can take those papers to court to defend yourself.”
While these tools can provide valuable support, Bowan cautioned that relying solely on AI carries risks, especially given the potential for inaccuracies, or a lack of contextual understanding in complex legal matters.
