Earlier this month, ILTA hosted London Insight, which was one of the largest legal tech events in the UK this year. There were plenty of great speakers to see, but for Artificial Lawyer one session in particular stood out: ‘Legal Innovation: More Than Just Artificial Intelligence’. This featured four speakers:
- Ben Gardner – Information Architect, Linklaters
- Jan Van Hoecke – CTO and Co-Founder, RAVN
- Drew Winlaw – COO and Co-Founder, Wavelength Law
- Milos Kresojevic – Innovation Architect, Freshfields
The following is an overview of some of the key points made during the session.
To begin the session each of the four speakers set out some of the key ways they approach the use of legal tech, especially AI. Ben Gardner kicked off with a description of how Linklaters helped clients with a simple, yet effective name checking system to conform to banking regulations.
Previously it had taken on average 12 minutes per customer to check that details were correct. Linklaters’ internal tech team made use of open source software to build an automated triage system that could sift through the textual data. This homemade solution hugely reduced the time involved.
Gardner pointed out that the key philosophy was: identify a pain point, develop a solution that could prove its value, then scale up the solution. He added that by using open source software they also avoided a high cost and hence reduced the risk of blowback from the law firm’s partners if the solution didn’t work.
Another Magic Circle tech expert, Milos Kresojevic from Freshfields, then gave an overview of some of their approaches to legal tech.
The firm’s innovation team had launched around two and a half years ago and had worked on using AI for due diligence and also predictive coding in e-discovery in the US. They had also worked closely with the firm’s process centre in Manchester, known as the global services centre, to integrate tech applications into the firm’s normal work routines.
Kresojevic added that they had also begun to explore the area of smart contracts and blockchain. He also wondered whether blockchain was going to disrupt AI (n.b. something we’ll come back to in the future). He added that Freshfields was also working with the Manchester University Computer Science School to improve on semantic search systems.
But, probably the most important point was that the firm had taken the decision not just to experiment with AI and new legal tech in isolation, but had sought to openly engage with a wide range of partners and groups across the legal tech ecosystem.
‘How do we create value for the client?’ asked Kresojevic. ‘It’s about creating a network effect in the value chain.’
This meant the firm getting involved with start-ups, conferences and hackathons, as well as working with vendors of legal AI systems. This was because by engaging with the entire ecosystem of new legal tech they were creating conditions and outcomes that would help the firm and their clients.
Artificial Lawyer has to say that this is a very commendable approach and is in stark contrast to some law firms that seek to operate ‘in secret’ in terms of legal tech for fear they might lose some perceived competitive advantage. Of course, in reality the competitive advantage comes from the benefits of engaging with the people developing AI and new legal tech applications. Great innovation rarely takes place in a vacuum.
Jan Van Hoecke of RAVN then provided an example of where the UK-based company’s AI engine was put to use. A client needed to examine thousands of documents for risks in relation to bank overdrafts. Van Hoecke and RAVN had to figure out how to triage the risk in the documents without having to do this manually.
The method used was to cluster documents and tag them relative to risk, then use a ‘deviation calculation’ to identify the most risky. The AI natural language processing (NLP) system could then zip through the thousands of documents and very quickly classify them. Van Hoecke summed up that this up was a classic example of a client gaining ‘efficiency and providing risk management’ by using an AI system.
Finally, Drew Winlaw of Wavelength Law added a different perspective by looking at legal tech in terms of what he termed ‘outcomes focused service design’.
He provided a flowchart for how lawyers need to think about legal tech, starting with the business’s corporate memory, i.e. the knowledge, documents, templates and other data that had been generated over the years. From there a business could then look at developing ‘playbooks’ that helped to steer legal processes, risk appetite and patterns of document creation.
The playbooks would also benefit from using tech-assisted analysis of the corporate memory, permitting clustering of data and tagging of key concepts and ideas.
This ‘industrialisation’ of business knowledge and then leveraging it to improve efficiency and reduce risk could then be followed up with automated contract construction, automated document review (which may use AI) and wider use of electronic signatures to speed the execution of matters.
The end result would be a holistic approach to legal tech, not looking for just one solution to one problem, but seeing it as an end-to-end integration of different types of legal tech across the business.
There then followed some questions both posed by other panel members as well as the audience. Freshfields’ Kresojevic talked a little about how the process of launching the internal innovation team had unfolded.
‘It was like running a start-up inside a law firm, but with a different agenda and pace [to the rest of the business],’ said Kresojevic.
He added that they had to think carefully about their resources, which were limited, and how best to seek out and discover what was going to be useful.
Interestingly he stressed that winning a hackathon last year for an access to justice application was far more important than just a bit of kudos for the firm. The win helped to engage partners at the firm, which in turn helped to provide more support for the innovation team to press ahead with their ideas.
In short, engaging externally in the legal tech world gave highly valuable credibility to what the innovation team was doing in the eyes of the partners, who although they may not have understood the underlying technology, certainly did understand and respond to winning prizes for innovation. That in turn then hopefully will lead to greater support and resources for cutting edge legal tech innovation at the firm.
RAVN’s Van Hoecke then looked at the challenge of getting law firm partners to engage with legal tech from the perspective of a vendor.
‘We had a lot of teaching to do. It really opened my eyes [when we went into law firms] as to how little they used technology,’ said Van Hoecke and stressed the role of tech companies in educating lawyers about what they did.
This was another important point, especially in relation to AI. Most legal tech has remained in the domain of law firm IT heads, who are not always listened to by the partners in the firm. With the new wave of legal tech, especially AI that actually does legal work, the tech companies need to appeal to the partners as much as, if not far more than, the IT staff inside law firms.
This of course is a challenge because some partners don’t know much about the new wave of technology such as AI, while tech companies are not designed to be educational centres for lawyers. Yet, this is where we are today.
Van Hoecke perhaps summed this all up by saying: ‘Partners need to buy into this technology and not just be a consumer of it.’
He concluded that this was a two-way street: ‘This is bi-directional. We are quite far as vendors from the problems of your clients. So, we are asking: what is the problem you are trying to solve?’
And that is perhaps a good place to leave this session, which it is fair say that over and above the examples of how to use AI and automation had one central message: engage with the new wave of legal tech companies and learn.
Thank you to ILTA for helping to further this cause and bridge the information gap between lawyers and the new wave of legal tech.
Be the first to comment