In February 2016, a London court supported the use of predictive coding software in a legal disclosure process, which often involves lawyers receiving huge volumes of documents from those representing the other side in a case.
In Pyrrho Investments v MWB Business Exchange, Master Paul Matthews of the Chancery division supported the use of software in scoring documents for relevance. He found there was no evidence that software would be less accurate than manual review and keyword searches. He added that software could provide greater consistency in searching more than 3 million documents that could be involved in the disclosure. A final reason was that both sides had agreed to the use of the software, which would be much cheaper than a manual search – they just wanted the court’s approval.
However, in May, the High Court went further when two undisclosed parties disagreed on whether predictive coding software should be used. The “petitioner” wanted to review inboxes using an agreed list of search terms and the “respondents” wanted to use software, on grounds of lower cost and better results. Partly relying on the earlier judgment, the court ordered the use of the software.
These decisions show how parts of the legal profession, often perceived as a technology laggard, are catching up and starting to surpass others in use of automation and artificial intelligence (AI). “Lawtech” is aiming to transform the profession in the way that fintech is changing finance, with a Legal Geek conference planned for this October in Shoreditch.
In January, the Law Society – which represents solicitors in England and Wales – published a report on technology innovation in legal services. Based on interviews with early technology adopters, it found that technology would have a “profound effect” on law firms’ staffing, pricing and location.
As last year’s court cases suggest, many corporate law firms are using software to search documents, according to the report’s author, Dr Tara Chittenden. But some are extending its use into managing legal spending, supply chain, documents and customer onboarding. “I think people are still fairly early on in exploring what it can do,” says Chittenden, with natural language processing looking like a particularly promising area.
She says other professions with more interactions with consumers may be ahead at the moment, but adds: “There is a stereotypical view out there that law is behind. I certainly don’t think that’s true. There’s a spectrum – some firms are really clued up and are really ahead of the game in what they’re doing, and I think others are struggling.” Law firms will also use software to monitor and reduce costs, and to gather more data on clients, with the eventual aim of predicting customers’ needs.
This may mean IT departments taking a lesser role. “The firms that I’ve seen which are doing better are those where IT decisions are integrated across the whole of the business,” says Chittenden. “The process of choosing and using an AI or an automation system has kind of bypassed the IT department.” The supplier is more likely to approach a head of innovation who will discuss it with senior partners, she says.
UK-headquartered law firm Pinsent Masons has built its own AI system, TermFrame. Developed by the firm’s head of research and development, Orlando Conetta, it guides lawyers through tasks by bringing in documents, templates and information on legal precedents, such as by allowing lawyers to assess legal risks against consistent, detailed criteria set centrally. “This means we are able to delve deeper into the requirements of our clients and scale up our work to a level that would hitherto been unfeasible,” says Conetta.
While AI and big data make processes such as document review more efficient, it is at its most effective when combined with human expertise. “The system may identify and classify a clause that pertains to transfer rights, but the model will be unable to confidently assess whether the particular clause actually prohibits or permits the transfer of rights and obligations between parties,” says Conetta. “A lawyer would need to read the clause and make that determination, while AI ensures that the lawyer can find the clause quickly and only needs to assess the provision once, across all the documents it may be found in.”
Conetta believes that using AI across an organisation’s legal work will have implications for heads of IT. “Traditionally, CIOs have been focused exclusively on the large challenges relevant to the operational aspects of running a law firm, such as document management, email, office productivity, and so on,” he says. “The advent of AI means that technology teams will effectively become another pillar of legal service delivery in the practice, and computer scientists will find themselves working alongside lawyers and partners on actual transactions and disputes.
“This will be exciting, but will bring with it new pressures to ensure that technology teams have a good appreciation of the respective legal subject matter. And, in parallel, legal teams will need to adapt to embrace new cross-disciplinary team structures.”
Read more about AI and the law
- What is an artificially intelligent attorney (AI attorney)?
- Law firms are rapidly realising that if they don’t latch onto the benefits IT can deliver, they will lose out to rivals.
- How legal health specialist Hempsons Solicitors rips out its existing Cisco infrastructure for increased functionality and cost savings.
Pinsent Masons already has scientists and engineers working with lawyers on categorising clauses and data within legal documents and structuring AI legal assessment questionnaires.
Suppliers to law firms are also aiming to provide automation. LexisNexis, which owns long-established legal publisher Butterworths, has developed automated tools based on its information. “We are trying to move the risk of automation projects from the client to ourselves,” says document automation manager Adrian Stafford.
The company offers automated documents for lawyers and non-legal staff, with the latter including sales and non-disclosure agreements, which can be populated by a questionnaire. Stafford says the former are designed to generate a good first draft of a document, ensuring that one piece of information appears consistently in hundreds of places in a document, with lawyers doing further checks. But for the non-legal documents, “they are looking for something that’s signature-ready”, he says. Such automation saves time, but consistency and the avoidance of error can be more important.
If law firms provide such specific versions of these documents to non-legal staff at their clients, questions including securing access to the system and the location of hosting must be considered. “The IT department has a role in all these issues,” says Stafford.
He says the use of automated documents has spread rapidly across law firms, including into their own processes, such as client engagement letters and human resources. “It’s trying to provide what could be called a digital butler that will sit alongside the lawyer as they work to be able to identify what it is they want to do,” says Stafford. “If it’s a simple task, be able to do it for them, and if it’s a more complex task, be able to offer them the advice and materials that they need to deliver it in the way and at the point at which they need them.”
Such automation will change what lawyers do, particularly as smart contracts are introduced that can carry out transactions as well as codify them. “What starts to become important is the intention of the parties, negotiating on intentions rather than necessarily the words on the page,” says Stafford. “The role of the lawyer will change to become much more of a legal business adviser to their clients, assisting them in what they can and can’t do, what they should and shouldn’t do, rather than just working out that there needs to be a comma at this point in this clause.”
The wording of documents will become more standardised, and automation and AI will focus on helping corporate lawyers in their roles as business analysts, he says.
Much of the progress on automating process is taking place at larger law firms. Legal startup Amiqus is targeting firms with a small number of offices. It carried out research to find out what they most wanted automating – and, as a result, has launched a service for anti-money-laundering checks, something firms will have to carry out from June this year.
“Law firms are a traditionally risk-averse, conservative profession – they’re not what you would call early adopters,” says Amiqus chief executive Callum Murray. He says the company originally planned to focus on larger firms, but adds: “Smaller law firms are actually at a greater risk, because they lack process automation at all.”
Amiqus plans to develop a range of services for smaller firms as well as having a platform with APIs that can be used by developers for larger firms. “We’re not trying to get rid of solicitors as such, we’re trying to future-proof the way they deliver their service through automating,” he says.
Automated contract analysis
The company is looking at automated contract analysis – already used by larger companies – and an online dispute resolution system that can help settle many cases more cheaply than in court. “The issue with that at the moment is access to the volume dataset,” says Murray, with a lot of data currently behind paywalls. The Ministry of Justice is looking at openly publishing some court information, he says.
“It’s going to take some time to change that, but once that changes, it’s really going to open the floodgates to an open innovation approach.”
This use of automation and AI could allow firms to consider potential clients by collecting data, then comparing it to previous cases. “You could triage cases to say is this a case worth taking on, what’s our chance of success – almost a forecasting tool for litigators,” says Murray. Another option would be a third-party system that collected information from possible litigants, then offered their cases to a number of firms, which would choose whether or not to offer their services. “Effectively, the lawyer becomes far more accessible because you don’t have the huge barrier of the discovery cost,” he adds.
Murray thinks CIOs will have to work more closely with legal officers to allow such automation. “It’s not entirely a technical problem – some of it is process, some of it is change management,” he says. But a more automated legal profession will need to work closely with technologists.
Law catching up
Murray believes law is catching up with other professions, most of which have already adapted to new technology, but that it is more likely that automation and AI will support lawyers rather than entirely replace them in the fashion of financial “robo-adviser” services. “People aren’t quite ready for a computer to say no,” he says. “They are used to that on a mortgage application, but when someone is telling you that the custody of your child is not going to happen, you want a bit more of an explanation.”
Richard Susskind co-developed the world’s first commercially available legal AI system in the 1980s and has recently co-written The Future of the Professions about the impact of technology on professional jobs. “In the 2020s, legal professionals will have a stark choice – compete with machines or build the machines,” he says. “By competing with machines, I have in mind that human lawyers will be doing things that machines cannot.
“By building the machines, I have in mind recognising that there will be AI solutions in the future for many of the problems that bring clients to their lawyers today; and so the way to meet clients’ needs will be to be involved in building these AI systems.”
Susskind says AI is likely to carry out work including due diligence reviews in transaction work, predictions of court decisions and online dispute resolution. “All of this means that legal professionals will not just be legal advisers, they will also be legal knowledge engineers, legal data scientists, legal technologists and legal process analysts.”
Although he reckons medicine and tax are more advanced, Susskind believes the legal profession has invested heavily in AI – and is ahead of teaching and the clergy.
Legal trends prove the power of data