Making AI work for legal professionals: Practical lessons from the experts
What is the transformative impact of generative AI on law and legal services? This is the key question Jorn Vanysacker, co-CEO and founder of Henchman, addressed together with Dom Conte, Chief Product Officer at Avail, during a co-hosted webinar. They paint a compelling picture of AI as a valuable sidekick, poised to revolutionize the legal industry. To make sure the proof is in the pudding, the two gentlemen are joined by Katherine Crowley, Practice Development Manager at the esteemed law firm Womble Bond Dickinson. She treats us to some insights from the field.
Time no longer means money
The pressure on lawyers is building up: clients who have to wait are turning to ChatGPT for a contract proposal; which raises questions about the added value of a lawyer. Customers want to know how lawyers are implementing tools to make their bills less expensive. Generative AI will help to save time, reduce costs, and bring efficiency gains to the legal profession, this is something both Jorn and Dom strongly agree on. They believe AI will help lawyers shift from routine tasks to high-value and innovative work. This shift could lead to a significant change in billing models, replacing the billable hour with a more value-based approach.
Artificial Intelligence will enable lawyers to assist clients in unprecedented ways by making data-driven decisions based on past work for example, and deliver legal services more effectively and efficiently than ever before.
Do legal professionals then have to become prompting experts? As Jorn puts it: “AI is something that has to work for the user, and not the other way around. It’s not something you should be setting up, but something that has to be available as a sidekick. In that way, I don’t think it’s going to replace the lawyer of tomorrow.”
Survival of the Tech-Savvy
According to Katherine Crowley, there is a unique momentum – if you look at the usage figures, no digital tool has ever raised as much interest from lawyers as Generative AI, with allegedly about 90% of lawyers having tested large language models like ChatGPT: “We need to take this interest and enable our lawyers so that they know both the law and how to work with AI and other technologies to get the fastest and the best outcomes for their clients”.
The experts even appear to be ardent supporters of corporate-level Darwinism: those who fail to adapt will eventually perish. AI won’t replace lawyers, but lawyers who use AI will replace lawyers who don’t.
“Law firms will get left behind very quickly if they don’t engage with generative AI. It is here to stay. We’re only seeing the first generation and even then it is super powerful. It’s amazing what it can do. I want lawyers to be fearful if they don’t engage, they need to be scared… Yes, we all need clever lawyers because AI will only get us thus far. And we definitely need lawyers who know the law. But let’s change who we’re looking for. We need people who are interested in technology.”
Katherine Crowley, Womble Bond Dickinson
In recent months, the market appears to have matured. “The questions we hear at Henchman have also evolved from wanting to do something with AI to specific use cases.”, Jorn says. This is partly due to pressure from clients to work more efficiently, but another driver is certainly the war for talent. If law firms want to be able to hire those people who are interested in technology, they need to be frontrunners in the technology they use too. A third element at play is the pile of data legal firms have stored: “Legal firms are sitting on so many contracts, so much data, often unstructured. AI and Large Language Models make it possible to turn that into something that you can use as a lever to enable your lawyers, your teams, but also your clients”, Jorn explains.
Tech alone is not enough
We’ve all seen the party tricks with AI where someone asked ChatGPT to come up with a poem or to answer a funny question. And generally speaking, it works very well. But tech in itself is not a solution. “Just throwing Generative AI at the problem won’t fix it”, Jorn states firmly. You need to look at what the pain point is and try to solve that. And the outcome might be powered by Generative AI, but that doesn’t necessarily have to be the case.
According to Katherine Crowley, the time is now for lawyers to engage and experiment with AI. Customers expect the legal sector – which is traditionally a slow sector when it comes to tech innovations – to not miss the boat. Some law firms are even rolling out their own rebranded versions of ChatGPT and getting a lot of press coverage with these projects, but if you ask Katherine, that is not the way to go: “I think you need an awful lot of resources to be able to do that. And the majority of law firms don’t have those resources or, more importantly, the expertise at the moment.” Her proposal? Look at what suppliers are doing and broaden your knowledge to decide which tools can be valuable for your firm.
Handle with care
When it comes to legal tech, AI needs to be treated with caution. It can do wonderful things, but of course, AI has its capabilities and limitations. As long as the vendor is transparent about what the tool is designed to do, how accurate it is, and what it can or can not do, they are empowering the lawyer to make data-driven decisions.
“It’s very easy to criticize AI for accuracy, but when you compare it to the output human beings provide, it becomes a lot easier to justify why and how you should be using AI”, Dom Conte explains. Humans get tired, get confused easily and sometimes even tend to make things up. It’s the synergy between the technology and the knowledge of a skilled lawyer that makes it a powerful combination.
“If you put a team of paralegals on a review, you would then have a more senior lawyer provide the last set of eyes. It’s exactly the same with generative AI. However, it’s so much cheaper. In the past, you had to pay both the paralegals and senior lawyers. Now, using generative AI it’s not only cheaper but more reliable as well.”
Katherine Crowley, Womble Bond Dickinson.
To wrap things up Jorn highlights one key challenge in generative AI output: you need to make sure that the output generated by AI doesn’t only use your tone of voice, but that it also aligns with your specific context. “It generated something that sounds like us, that has our tone of voice. But is this something I would use in this particular case? Is this something the senior partner for whom I am drafting this contract or clause would use?”, Jorn rightfully wonders. “And this is something that large language models will never be really good at, because they are effectively very good at predicting the best next words, but struggle to match it to a particular setting or environment”.
“It generated something that sounds like us, that has our tone of voice. But is this something I would use in this particular case? Is this something the senior partner for whom I am drafting this contract or clause would use?”
Jorn Vanysacker, Henchman
The billable hour is under pressure.
It will become increasingly hard to stay ahead of the race by only billing by the hour. However, moving away from this way of working is not apparent: billing by the hour ensures no surprises for the lawyers, in case the scope changes. For the clients, however, the billable hour feels uncomfortable and sometimes like writing a blank cheque. They don’t necessarily want it dirt cheap, they want predictability and budget forecasting.
We’ve asked eight legal experts where they stand. And, interestingly enough, we got just as many different opinions.