Law firms can choose to either reject or embrace artificial intelligence – but ready or not, the tech revolution has arrived. Story by Felicity Nelson and Stefanie Garber.
Automation is an idea that either sparks genuine excitement or deep fear in lawyers. Having a computer extract the menial, repetitive tasks from a lawyer’s daily grind is a satisfying idea. But a future dominated by artificial intelligence (AI) – where lawyers’ analytical expertise is outstripped by the machines around them – is a troubling one.
When doomsday sayers suggest that ‘robot lawyers’ are only decades away from taking over, the feeling in the legal industry is, quite rightly, one of disbelief.
This may explain why so far there has been a snail-like uptake of AI technologies by law firms in Australia and around the world. As legal consultant Andrew Baker, a director at Janders Dean, puts it: “To be honest, globally, there is a lot of AI chatter, but the impact at present is rather weak.”
But, while the enthusiasm is lacking in the legal world, other industries are rapidly integrating AI. A recent report by Bank of America Merrill Lynch showed that 47 per cent of American jobs are likely to be automated by 2020.
The staggering gains that can be made by augmenting lawyers’ work with AI make change almost inevitable, according to consultant Dr George Beaton.
“Anything that’s faster, better, cheaper, has got to succeed.”
As AI sweeps through the industry, cutting costs and reducing labour, traditional law firm services and client relationships will be reinvented – or even replaced.
The future is now
While adaptation has been slow so far, some of the technologies that will upend the legal industry are already here, according to Julian Uebergang, managing director, Asia-Pacific at Neota Logic.
“This is not something that is coming in the future, this is something that is happening in the here and now,” he says.
AI takes a number of forms in the law, with some technologies closely mimicking human capabilities through machine learning, while others use data mining and analytics, bake expert knowledge into logical systems or use software to automate contract and document creation.
Machine learning was made famous through charismatic supercomputers, including Deep Blue, which beat a human at chess in 1996, and Watson, which demolished its opponents in Jeopardy in 2011. These may seem like high-tech party tricks, but companies have quickly realised the value of leveraging machine learning in the legal industry.
Take IBM’s supercomputer ROSS, for instance. ROSS can read and analyse over one billion legal documents a second. It can process complex legal questions and respond with an intelligent answer in plain English, along with relevant citations. ROSS is a digital associate that doesn’t take coffee breaks. It can whirl away 24 hours a day. It is across the latest legislation and the more lawyers use it, the more it learns.
Data analytics AI has also hit the legal market in recent years. US company Lex Machina, for instance, has developed databases and software that can provide an accurate opinion about the odds of winning a patent suit. According to Dr Beaton, Lex Machina can achieve in “two minutes for [a few] dollars” what a team of five lawyers can with a budget of around $300,000 over three months.
Expert systems are a fertile area for lawyers as they allow for the outsourcing of repetitive tasks to computers. One company working in this field is Neota Logic, a US-based start-up that expanded into the Australian market just this year. Neota Logic provides a platform upon which lawyers can build ‘logic trees’ that divide legal problems into a series of simple online steps that can be easily followed by a client.
Basic AI services such as document automation have been around for decades, but are now picking up pace. The US has a number of start-ups battling in this growing market, including the behemoth LegalZoom and Google-backed Rocket Lawyer.
These businesses give clients direct access to document creation, removing the need for face-toface contact with a lawyer in some circumstances. Only a few start-ups in Australia work in this area, such as LawPath and LawAdvisor.
While LawPath’s document automation services do not involve the complex reasoning typically associated with AI, they do offer similar savings in time and money. According to Dr Beaton, document automation cuts the amount of time it takes to draft a contract from days to minutes, with a comparable drop in cost.
Given that these technologies are already on the horizon, lawyers are beginning to question what the practice of the future might look like – and how they will fit in.
The heart of practice
Lawyers spend five years-plus at university to qualify, followed by years working their way up the firm hierarchy. The thought of all the knowledge and experience being accessible at the click of a button has some lawyers spooked.
Yet Mr Baker believes a human element will always be necessary for the provision of legal services.
“I think the analysis and the work that a lawyer does to move a matter along still have to happen,” he says.
“Many of the deep, analytical parts of practice will still be done by a person, without question,” he adds.
In fact, one argument holds that expert systems technology will free lawyers’ time up to focus on high-level legal work where their analysis and knowledge can add the most value, according to Dr Beaton.
He pointed to document discovery in litigation – once the scourge of junior lawyers – as an area that already has been 95 per cent replaced with technology.
For lawyers who specialise in process-driven or bulk work, what this means is an open question. Sole practitioners and large national firms with large conveyancing or estates practices may find their contribution to day-today legal work significantly altered.
Even if human beings remain the primary producers of legal work, how those humans approach a problem may begin to shift.
Lex Machina’s data analytics program, for example, can reveal the prior behaviour of the defendant and predict outcomes in patent cases. Suddenly, lawyers can form data-driven predictions on the likelihood of a given strategy’s success.
More advanced AI programs may enable lawyers to get an ‘answer’ to a legal problem from a technology platform. Even in these cases, however, Dr Beaton believes the lawyer will remain in charge, able to decide whether to accept or reject that opinion – “It’s up to the lawyer to make the call,” he says.
This could introduce an element of ‘best practice’ to the law, which Mr Beaton likens to a medical approach: gathering information, making a diagnosis and then presenting a solution.
Underpinning all these trends is data and the sudden availability of large swathes of information. According to Mr Baker, law firms are likely to see a new kid on the block, hiring data experts – possibly even non-practising experts – who specialise in breaking down the available information.
However, in his view, even traditional lawyers will need to get their heads around data.
“While day-to-day practitioners are not going to be looked for true data analysis, they will understand [why] data is being collected and what it can help prove or predict,” he says.
“Clients are going to look for more and more data from their service providers and partners to mitigate risk and plan for the future,” he says.
Lawyers might also take on new roles behind the computer screen. The Neota Logic system casts lawyers in the role of process designer, drawing on their legal expertise to engineer automated question trees, and then granting the lawyers IP rights to each system they build.
“I think there is always a need for the systems themselves to be maintained by experts,” Mr Uebergang says.
“Because of the changing nature of what we do in terms of sciences and technology evolution, there is always going to be work for lawyers to have to tackle and things that need the law to be applied.”
Learning to relate
While human lawyers will still be necessary in the future, how they deliver services to clients could shift dramatically.
Currently, lawyer-client relations tend to be primarily ‘analog interactions’ – phone calls or meetings. At most, lawyers will send clients an email, but technology has been slow to intrude on this realm, Mr Baker believes.
In future, he predicts firms will need to provide a greater number of digital connection points for clients. In fact, service delivery could be where the greatest changes occur for law firms and lawyers – and the pressure will come from clients themselves.
“Changing the way lawyers behave is hard, and massive change hits cultural walls and slows adoption,” Mr Baker says.
“That’s why the buyer is so key to the equation. If the buyer is encouraging or demanding these approaches, the lawyers will take action more seriously. Without a ‘pull’ from clients, growth will continue but won’t reach maximum velocity.”
This new model could take a number of formats, from digital platforms to AI systems allowing clients to get quick legal answers online.
In some cases, lawyers will be removed from the equation entirely, according to Mr Beaton. In the US employment law space, for example, large-scale employers have access to apps produced by law firms that deliver advice on basic employment issues.
“The managers in these businesses can get instant advice about the situation anytime, anywhere, 24/7,” he says. “And if at the end of the advice, they still feel they want a lawyer, then they just press the button and call one.”
For clients looking for help on low-risk matters, this advice is “as good as the advice of a lawyer”, Mr Beaton believes. Indeed, it might even be better, given clients can access legal expertise for a fraction of the current cost.
While lawyers may feel displaced, Mr Uebergang believes these types of applications can deepen, rather than strain, the client relationship.
“Firms can expose systems publicly via their website or make it available to clients to create a level of stickiness with that client and solve particular problems that are related to those claims,” he suggests.
One concern is whether automated legal systems are providing legal advice, which raises the questions: whose liability is it if a machine gets it wrong? According to Mr Uebergang, however, advice delivered via tech does not require special treatment.
“Often if it is not specific legal advice; they make that clear in the terms and conditions that are put forward in the system. Ultimately the experts themselves or the firm who is sponsoring the applications take responsibility for what the application does,” he said. “All we do is provide the platform.”
Another change could be seen in how clients select legal representation – and this could strike right into the heart of major corporate firms. According to Owen Byrd, the general counsel of Lex Machina, in-house counsel are increasingly using data analytics to choose their lawyers. Lex Machina, for example, gives clients access to law firms’ history with patent issues, the parties and the outcomes achieved.
“Having objective data about a firm under consideration can ensure that the selection is not made merely out of habit, but is driven by efficiencies and effectiveness, so you get the best, most qualified outside counsel for the job,” Mr Byrd says.
Lawyers may also be prepared to embrace novel billing methods – for example, a fixed rate fee for a given process or even a pay-per-use app, Mr Uebergang suggests. While fixed fee approaches are not new, technology allows firms to automate the process so that each service delivered costs the same amount.
Ultimately, however, Mr Baker believes lawyers should not fear losing their clients to machines. In his view, these trends will lead to increased trust between firms and clients.
“There is going to be much more transparency, both on the billing side but also on the practice side,” he said. “That technology wrapper is going to change how firms and clients interact with each other on a pretty broad level.”
Behind the scenes
What happens in the back-office – away from the client suites and conference rooms – determines how efficiently the firm functions as a unit.
Like any business, firms have always been seeking greater efficiency, better co-operation and more bang for each buck. Exploring AI technologies is the same process that led firms to replace typewriters with computers, or message boys with phone systems.
“If you go back 10 or 15 years, there was a lot of investment in setting up document precedents and things internally within law firms to streamline the creation of documents,” Mr Uebergang says. “That largely stemmed from what was lower-value work from the client’s perspective in terms of what clients are willing to pay for. I think this is just an extension of that.”
In his view, any process involving repetitive questioning lends itself to being automated.
Mr Baker believes the approach to practice management – particularly how services are orchestrated and sequences – will mature in coming years.
“This revolves around the convocation, analysis and synthesis of the legal processes,” he says.
The overarching goal is examining how legal services should be designed and delivered.
“[It’s about] trying to figure out if we’re giving engagement or giving time; how much of those should be deployed on a matter to optimise it; and how to enhance the client experience,” he said.
The linchpin, in his view, is data – and he predicts there will be a surge of legal activity in this area.
This philosophy also underpins the Lex Machina model. Mr Byrd believes firms will begin to draw more heavily on data to determine which work has the most value. In a world where data is accessible on every device, it is inevitable that data will begin to shape law firm strategy.
While predicting the future of the legal industry is difficult, if not impossible, experts are fond of trying.
“The rate of change, and the growth of cognitive computing is so vast that nobody really knows how this is going to pan out,” admits Dr Beaton.
“There are people like [Professor Richard] Susskind who say there will be a lot of replacement, and there are people who say there won’t be much at all.”
One thing that most agree on is that AI has the capacity to modify the ways that clients reach lawyers. AI could also have a levelling effect, with smaller firms having a bigger chance to compete, according to Mr Byrd.
Lawyers also question whether technology will decrease the market for legal advice, but Mr Uebergang believes these fears are unfounded.
“As the advice they are giving becomes more affordable, we think more people will utilise that advice and go to lawyers more often,” he says.
Legal aid and pro bono lawyers also have much to gain from the increases in efficiency that AI promises, according to Mr Uebergang.
“These organisations are trying to get to as many people as possible,” he says.
“Often they just don’t have the bandwidth to deal with the number of inquiries that come in.”
While some law firms are putting a toe in the AI waters, there is little sense of urgency from practitioners. Mr Baker suggests firms are keen to gain any competitive advantage but are somewhat baffled by AI, with many “having a hard time working out how to deploy it”.
“Firms are good ‘problem-solvers’, but this is more of an exercise in ‘problem finding’ or ‘opportunity finding’,” he says.
“Fear of the unknown” is one of the factors holding firms back, according to Mr Uebergang. “There is an element of fear associated with change in any industry,” he says.
Perhaps the most “pragmatic view” for law firms, according to Dr Beaton, is to adopt new technologies as they come through.
“Because if you don’t get into the augmentation route, then it’ll be like the [person] that didn’t adopt email and was left with the postal system or didn’t get into word processors, or didn’t get into document management systems,” he says.
“Whether it be barristers, solicitors in private practice, in-house lawyers [or] court systems, [you should make sure] you’re equipped with these new tools.”
Like this story? Read more: