Legal AI Forum – Building a GenAI community 

The Legal AI Forum took place on Friday 11 July in London led by NetLaw Media and legal AI provider Jylo, bringing together senior law firm leaders to focus on the business implications of the latest advances in AI.  

The legal tech events calendar is jam-packed these days and there’s more competition than ever for people’s time. With this event, the organisers focused heavily on workshops, use cases, and discussion, which made for a thought-provoking day. Interestingly, they divided the streams between in-house and law firms, meaning there were some senior corporate counsel among the attendees. 

NetLaw Media runs the British Legal Technology Forum and there were strong overtones of BLTF at the forum, which was chaired by Professor Richard Susskind. Other key speakers included Christina Blacklaws, Jenifer Swallow, and Christian Toon, who host the main stages at BLTF. 

However, the forum is heavily focused on content, knowledge exchange and community, which is a theme after Legal IT Insider’s own heart as we prepare for our 30th anniversary in October. 

We are at the foothills of GenAI 

Opening up the day, Susskind, a legal tech author and public speaker who doesn’t really require introduction, said that in his opinion, we are very much at the foothills of generative AI. “We’re at a very early stage and it’s important to gather and speak openly together,” he said, adding of the forum: “This is about building a community.” The UK has led the world when it comes to law, Susskind said. Now is our opportunity to do the same thing in the application of AI to the practice of law and administration of justice. 

Susskind reiterated that there is a big distinction between ‘short-term’ and ‘long-term’ AI. Short-term involves overstated predictions and reactive service. Long-term predictions understate the reality, and Susskind said that he had no doubt that as we move to the end of the 20s and 30s, AI will revolutionise legal practice. 

In terms of what will change between now and then, the answer could be everything, but one important prediction is that we will find ways of overcoming hallucination and bias. It was amusing to hear Susskind say: “As with all tech that starts out a bit dodgy, it’s reasonable to think that these systems will become better.”  

In the long term, Susskind said, we will get to artificial general intelligence, where AI systems match us across the full range of our cognitive ability.  

Short term it’s about equipping lawyers with tools. Longer term, it will be about empowering clients to do much of the work that currently requires specialist advice. 

Susskind continued to set out his vision in some depth as to the difference between short and long-term outcomes. Perhaps his most interesting reminder was that we need to think bigger: we will likely move from reactive to proactive law, with a fence at the top of the cliff instead of an ambulance at the bottom. Law will be conducted in an entirely different way. All of this will require bold leadership indeed – of a kind that we have not seen yet. 

To predict the future, we must understand the past 

Speaking after Susskind at Legal AI Forum was distinguished legal professor Michael Hoeflich, a historian of law and technology, who is a professor at the University of Kansas and has spent time studying at Cambridge. 

Hoeflich, who became interested in legal tech while at Cambridge, spoke about the five phases of technological adoption and how tracing some of the biggest disruptions of the past can help us understand adoption in the future. 

There are five phases to every technological disruption: Introduction; innovation; adaptation; disruption and transformation. Each phase is dependent on the one before in order to succeed – meaning that no technology can become transformative without causing disruption. 

Currently words like ‘innovation’, ‘ecosystem’ and ‘disruption’ have lost their luster, Hoeflich says, explaining that a population can show strong resistance to invention. Success is not dependent on those users. 

Hoeflich gave an example of another technical revolution that faced resistance – the introduction of the typewriter. This was strongly resisted by Scriveners, who oversaw document production and reproduction. Hoeflich gave a fascinating insight into the slightly toxic culture surrounding the all-male scrivener workforce. Remington, a major manufacturer of of typewriters, took a gamble in marketing typewriters as ‘women’s work’ and in less than 10 years, women typists were replacing scriveners. 

You can read more on that in Professor Hoeflich’s paper here – https://josephhollander.com/news-blog/five-phases-of-technological-adoption/ 

Stories from industry – The AI practitioner 

 The day was full of sessions from speakers talking about how they are using GenAI in their workplace. In a panel focused on ‘what it’s like on the inside’ for the AI practitioner, Alison Abell, legal transformation manager at Vodafone said that everyone in legal has Copilot and the team is rolling out a legal AI assistant (as yet undisclosed) to many. Vodafone is using GenAI in discovery to surface fraud risk, for example. Abell warned: “You can’t just buy an AI tool and expect magic. The key to success is having stakeholder engagement and building on that,” adding: “It’s more about having an open mindset and leveraging the enthusiasm of early adopters.” 

At Kennedys, Catherine Goodman is chief knowledge officer and AI lead. She is working with the senior leadership team to rethink junior training. AI, often presented as a disaster for junior training, taking away, as it does, the monotonous and repetitive work of the past, is an opportunity to work smarter and train junior lawyers better, Goodman said. 

The firm is putting in a new ‘cradle to grave’ competency framework, and a training programme that maps competencies. “AI robbing juniors of learning is the wrong statement,” said Goodman. “We need AI exercises to accelerate exposure.” 

Speaking on the same panel, Joe Cohen, who is director of advanced client solutions at Charles Russell Speechlys, discussed how CRS has a multi-prong innovation strategy that sees legal tech managers embedded in the firm’s practice areas. There is a strong focus on legal process improvement and one of the pillars is software development: they are building their own SaaS platform with a view to selling software instead of just time. 

In terms of training, Cohen said that where managers have previously not had sufficient time to train juniors, AI does, commenting that the firm is developing an AI tutor that can give multiple choice tests and provide experiential learning. “It will be richer training that what is lost, and they will learn more,” Cohen said. 

The focus of GenAI at CRS is pre-baked use cases and prompts and they “haven’t yet got to agents,” Cohen said. 

Last up on the law firm AI practitioner panel was Shilpa Bhandarkar, former partner and head of client tech and AI at Linklaters (now a consultant), who talked through the genesis of Linklaters’ AI strategy, how it has set the firm up for the future, and the importance of a structured response. 

When GenAI first exploded into the tech scene, Bhandarkar said: “We took two months out with the Steering Committee and came up with four scenarios: what clients would let us do; what adoption would look like; what tech would look like; and what we thought our competitors would do. We mapped out those four scenarios and had to agree where they would be in 18-24 months and then came up with an 18-24 month plan.” 

The team met every two weeks to benchmark and, if need be, rejig the plan. “Linklaters has always been a very structured firm and that set us up for everything from how to train people to where and who to hire and the best way to get very senior people to make changes.” 

Conclusion 

It’s always a positive thing if there’s too much content to write about, and that is the case here.  

There were fantastic presentations from Emma Dowden, COO at Burges Salmon, and panels including Bruna Pellicci, chief technology and data officer at Linklaters; Catriona Wolfenden, partner and product & innovation officer at Weightmans; and John Hunter, CIO of Council of Europe. 

The latter panel looked at understanding human behaviour, and some of the challenges and opportunities of GenAI. Chaired breakout sessions involved in-depth, scenario-based discussions around delivering your AI strategy. 

The numbers at the conference were small, which for attendees, was one of the appeals. 

Abby Ewen, COO at Browne Jacobson told me after the conference: “I liked the fact that it was comparatively small.  This meant that the breakout sessions were quite small too and this allowed them to be very interactive and participative.  Although there were separate stages geared around private practice and in-house, you had a choice of which sessions you wanted to do, and it was really good that the breakout sessions were attended by private practice and in house representatives, which allowed for some different perspectives.” 

All in all, this was a conference where you could really learn something. Susskind gave a keynote speech at the end of the day, commenting: “People treat AI like a football team: they are either for it or not. In reality there are opportunities and threats at the same time.” It’s in conversations that we can start to work out the nuances.