Speakers at the session included Elevate general manager of AI and data consulting Mollie Nichols, Herbert Smith Freehills director of generative AI Susannah Wilkinson, and Teneo director Janette Potgieter.

The session was moderated by Relativity discovery counsel and legal education director David Horrigan, who said that AI has “been a fundamental building block in how people are using technology” over the last two years.

Following the rise of ChatGPT and the “democratisation of GenAI”, this tech is “capturing people’s attention”, according to Wilkinson.

“When we look at work processes and systems, I think we’re now expecting more from technology than we previously were. But GenAI is not always the solution in isolation. And so, I think this is really interesting that [the report] shows, yes, I want to get rid of all the boring things that I do, but it’s just not quite working in isolation,” she said.

“It’s accessible to everyone, but that doesn’t mean that the right tools are accessible to everyone, and it doesn’t mean that the right tools are easy to embed or to deploy safely and responsibly. So, what we’re seeing is it’s hugely complicated in this area. We’ve got over 16 different GenAI tools that we’re currently deploying.

VIEW ALL

“There’s a huge time and energy that goes along with that for many of our clients. And what we’re seeing is that we can help, we can deliver the solutions that use the best combination of tech, including GenAI, much faster than the in-house teams of our clients can access the technology and test.”

Small organisations can compete with larger organisations, and technological tools help them do that, but data from the IDC study, according to Horrigan, shows that larger organisations are using it much more than smaller ones.

Potgieter said this is primarily due to resources as well as bigger companies having boards to drive forward more strategic decisions in terms of innovation.

“If you think about resources and infrastructure, corporations will have more people, more money, more technical infrastructure ready in place where they can run pilots, they can test the data, they can build their models, they can bring in external third parties, they’ve got a lot of resources available to them on that level.

“If you think about internal governance and risk management, again, they have existing legal compliance and data privacy teams in place that can pick up the nuances and navigate complexities around regulated environments and ethical complexities just in general,” she said.

“So, they have that advantage. And then also bigger corporations have got that strategic focus where the board’s really excited about innovation in general. So, it is an agenda and there is a framework in place, and there are timelines in place that they have to strive towards and action. Having said that, though, we do see a lot of small organisations and small law firms that are already using AI in very clever ways.”

Additionally, Horrigan noted that the study data also showed that paralegals and legal operations professionals use AI more – with lawyers and then IT being “the biggest non-users”, with additional surveys and reports showing that “paralegals tend to have greater technical skills” than practitioners.

Historically, Nichols said this wasn’t necessarily a surprise – and that for lawyers, GenAI adoption has been slower than technology-assisted review (TAR) was previously.

“If you look back in the day with e-discovery, it was the paralegals who ended up adopting the technology and then they would morph their careers into legal technology professionals. So, I’m not surprised that today, we’re still seeing that it’s the paralegals that are really using the technology and taking it to that next level.

“We grew up using technology-assisted review, understanding what machine learning and analytics would do to the document review process. For whatever reason, lawyers were quick to adopt it. And I think it was because of that learning process on what needs to happen in order for a document review to be defensible to ensure that the relevant documents are being produced, even if they’re not the attorney’s eyes on all of the documents. The phase that we’re going through right now, it’s more like a proof of concept,” she said.

“With GenAI, everybody wants to try it, they’re very eager to do that, but they also want to ensure that they evaluate the technology, they validate the results that they’re getting and not just accept them as is. And it’s going through that process to compare it to a TAR process. So, yes, they’re using it. But [in the next 12 months], I think we’re going to see a huge amount of adoption of GenAI because they’ve gone through the evaluation process, looked at the results and compared it to a TAR process.”

While the profession has already seen a “steep increase” in the capabilities of GenAI and large language models (LLMs), Wilkinson added that it would be interesting to see where this technology – and the profession – is in the next year.

“I think the way that the application layer is going to evolve in the next 12 months will be really interesting. We’ll get a lot more sophistication there and a lot more ease of use for the humans. And then, obviously, we’re going to see a massive increase in the way that the human users understand the technology, interact with the systems, and deploy that interrogation. I’m really interested to see this in 12 months’ time,” she said.

“What can the technology do really well? We know that we can point it to a limited set of information and it can extract information with a relatively high degree of accuracy and confidence. The flip side is, can we really use GenAI to draft documents? And are we there yet with the technology? Understanding what the LLMs are really, really good for, what the application layers are optimised for. So, [seeing] how much sophistication can we get from some of the other use cases will be really, really interesting in terms of broad adoption across legal.”