The legal landscape is changing with in-house teams and private practice firms alike seeking to transform ways of working through use of technology and well-designed process. We share our insights of how technology and design is being used in practice, the success stories and the pitfalls which can result in the impact delivered falling short of expectations. Plus we look ahead to future technology developments which could change the nature of the relationship between in house legal functions and external providers of legal services.
So to set the scene, there is a lot happening. Since 2013 and the first commercial applications of IBM's Watson software system, artificial intelligence is increasingly being applied to critical areas such as health, education, energy and the environment.
The key to understanding Watson and similar AI systems is that they are not word search systems, but instead are 'question answering' systems. They scan potential answer sources and evidence sources to learn and then answer questions posed in natural language.
Although it is unlikely that machines will attain intelligence that is broadly applicable to ours in the next 20 years, they will match or exceed human performance in many specific tasks. Watson software is already being used to provide second opinion on the best treatments for some US cancer patients,[i] and there are over 800 other organisations worldwide that have expressed an interest in creating apps based on the Watson platform.[ii]
That's a lot of progress since 2013, and technological change will continue and will dominate our future, both as human beings, but also as people involved in the legal services sector.
The global legal market has been estimated at USD $600 billion.[iii] That's an attractive market for technology providers and alternative delivery services to seek to shape and the products they are beginning to supply are becoming increasingly necessary to law providers. With that big a prize, it is unlikely that the legal the sector will be left to its own devices.
The US spend on legal tech is USD $3.5 billion,[iv] and automation could remove 100,000+ jobs in next 20 years.[v] McKinsey estimate that 22% of a lawyer’s job and 35% of a paralegal’s job can be automated[vi] as Machine learning and automation could replace many standard activities carried out by lawyers.
The principles used by 'robot' lawyers are the same as their human counterparts; they amalgamate data on a particular legal issue and advise on the basis of that collated information.
The legal system generates a huge amount of data. Judgements, precedents, interpretations of legislation all create more data and amongst it all will be facts and insights that could help win legal argument or be the key to unlocking a particular transaction.
Amongst the high volume of data that is available, machines are better at finding the relevant information and are more accurate. In an experiment conducted by LawGeex reviewing non-disclosure agreements with deliberate risks inserted into them - artificial intelligence tools finished with an average accuracy rates of 94% (the highest score being 100%) with the human lawyers scoring an average of 85% (the highest score being 97%), and whilst that accuracy gap is relatively (and reassuringly) narrow, the difference in the time taken was not. The human lawyers took an average of 92 minutes to review the NDAs, whilst the machine took just 26 seconds.[vii]
Until recently there has been little innovation in the way that the legal profession harnesses the data that the system creates, but that is now changing and we expect the rate of change to pick up pace in the near future.
Watson-like 'question answering' and machine learning systems now power "bots" that will dominate customer service of the future. Interestingly, Facebook abandoned an experiment after two AI programs appeared to be chatting together in a version of English that they invented which made it easier for them to work, but that no one else fully understood.[viii] Experiments such as these seem the perfect foundation for bots learning 'legalese' and having legal conversations.
Sceptics may say that people have been forecasting the advent of legal technology for many years but yet the substance of doing work in the legal sector hasn't changed. However to underestimate the impact of future technology, as a result of ones experience to date, is to ignore a fundamental aspect of technological development – and that is that progress is exponential and not linear. As one futurist puts it, change has a habit of beginning "almost imperceptibly and then explodes with unexpected fury".[ix]
In many areas technological change is beginning to explode. Computer processing power has doubled every two years for decades now.[x] Similarly the volume of business data worldwide doubles every 1.2 years[xi] and the magnetic data storage and internet bandwidth capacity keeps pace with that growth.
Also, the more people use a particular service, makes that service more valuable to them. As the user base of some of the software systems inspires enhancements and development, so those systems become increasingly capable and useful.
If we judge these and other technologies by what they can do now, fail to engage, or adopt a wait and see strategy, we risk being engulfed by the wave rather than surfing it.
So we are likely to live through some of the biggest technological shifts that will affect every aspect of our lives including our work.
Turning to the question of how technology may impact the work of in house legal teams, and firms like ours, we have identified at least three potential impacts. There could be others, but these are the ones that occurred to us:
1. Firstly, there are emerging and existing technologies being adopted to facilitate delivery of work across professions and industries that will have legal implications that we will need to know about and adapt to.
- An early example of this is people data. As systems operated by providers of goods and services have evolved to carry ever more personal data, so the regulatory frameworks protecting the privacy rights of individuals have grown, but there are many others.
- The use of Building Information Modelling Systems (BIMS) or drones in construction projects, and arising disputes.
- Blockchain, where the regulatory framework governing use has yet to be written
- Driverless cars and liability.
As technology explodes there will be more of these intersections where regulators and lawyers will need to codify responsibilities and commercial and personal rights. Our in-house counsel spend 50% of their time dealing with the impacts of technology.
2. The second potential impact is that technologies will continue to gain capability and the potential to make our work quicker, cheaper and potentially more consistent.
3. The third potential impact of technology, is that it may in time replace some of what lawyers do, may change some jobs, and may make some roles unnecessary altogether. Watson-like questioning answer systems coupled with bots may deal with all routine questions about what law applies, or what is the best course of action in a given set of circumstances.
To give some examples of tasks where technology is already changing the nature of the work undertaken by legal advisors:
Previously when it came to disclosure, a trainee would be tasked with creating a list of documents and diligently cross referencing it to a shelf of files. Once the documents were indexed, production of court bundles might have involved many hours at a photocopier.
However, with the advent of e-discovery tools, we can sift through hundreds of thousands of documents in the same time, can generate indexes and make the relevant documents available to the other side without having to go anywhere near a photocopier.
We also have specialist document review software, which searches for provisions in contracts, particularly useful in due diligence exercises. Many commentators have observed that document review and extraction technology capabilities will do for corporate transactions what e-disclosure has done for litigation.
Document review software operates, not by searching for words, but by learning concepts and the more you use it, the more it learns and the more accurate it becomes. Although even with it's out of the box functionality, it can provide a 40%-50% efficiency saving on human time for equivalent work.[xii]
These types of tools each started in a particular sphere, but are quickly expanding their capabilities, as once the information has been captured, it can be used in a myriad of different ways.
For example, e-discovery tools can help you to better navigate through the information after the litigation has ended. This same software combined with text analytics can also help identify the 'emotion' of a document. This could help uncover useful insights, eg. If e-discovery tools can isolate documents with an aggressive tone, this may enable us to get to the nub of the case quickly, or even find the 'smoking gun' email on which the case hinges.
There are lots of different document automation tools available. These enable lawyers to produce contracts or other documents through the use of a questionnaire format. The output can then be used as a first draft or be tailored for a specific purpose.
To a specific example, we've used automation to produce thousands of supplemental GDPR agreements on behalf of clients, and, at first, our clients were sceptical. This was because we didn't produce any agreements for almost a month and the client was understandably nervous with the deadline fast approaching, but it was during this time that we were building the automated template with the variables that the client required and organising the client information into a useable format. As soon as this was finished and approved, we were able to produce all the required agreements in seconds. So technology may also disrupt expectation of what progress looks like.
The current shift is not only towards automation, but also giving clients self-serve options to generate their own documents; this is not without its risks. Consider whether you would be happy for your business to produce their own legal documents without any direct input from the legal team or appointed panel? Determining who can use such tools, and for what types of contracts they can be used for, is potentially a role that will fall to in -house lawyers in the future.
Information capture, data transfer and data analytics
With collaboration platforms and online portals providing easy access to documents or key information now in wide use, constructive collaboration between legal services providers and between clients and advisors is much more possible. Whilst there may be concerns about sharing information, which cannot be underestimated especially since the introduction of GDPR and a number of high profile failures by big companies to keep confidential information confidential, businesses that collaborate tend to be more successful and likewise business that collaborate with their legal functions also tend to get better outcomes.
Capturing information in a central place also makes it possible to overlay data analytics and produce better reports on which to base future decisions.
The examples above still require experienced lawyer input, but there are also other technologies emerging which will in due course replace lawyers.
- Chat bots, already heavily used in customer service, they are making inroads into the legal services market. Currently the reported most successful chat bot is an app called DoNotPay which in the US saved motorists a total of $13 million in parking fines. Chatbots will be able to guide you through processes and recommend the best next action [xiii]
- Predicative analytics, information captured (perhaps through use of other technologies) can also be used to predict outcomes. In one experiment, a system called CaseCruncher was given the basic facts of hundreds of PPI claims and asked to predict whether the Financial Ombudsman would allow the claim. CaseCrucher predicted the outcome corrected in 86.6% of the claims.[xiv] If you are involved in litigation, you will know that lawyers rarely like to give a view on prospects of success much above 65%, so technology that can predict outcomes to that level of certainty is potentially a game-changer
- IBM Watson, a software tool that is currently used in the medical industry, but is being taken to 'law school' by a company called Ross Intelligence. The application could allow you to ask your questions in plain English and then search the entire body of law to return an answer[xv]
- Blockchain and its ability to guarantee provenance is of use to lawyers and this is being explored by Land Registry – as a way to synchronise and optimise participation in the property market.[xvi]
Why is it relevant to you?
If technology is changing the way your firms deliver services, engage with customers, collaborate in the supply chain or indeed impacts any other part of your operations, there will be legal questions arising.
In the first industrial revolution, there was initially no regulatory framework to govern innovation, imitation and competition. But as the means of production grew exponentially, as more suppliers brought more goods to more consumers, an ethical international and treaty-based regulatory framework emerged to deliver fair competition and protection of intellectual property. Before that it was a bit like the Wild West. Interestingly, it was then an emerging industrial nation – the US- that was then arch infringer of IP, in the same way that it is emerging superpowers today who are jumpstarting their own development in the same way.
As exponential development of technology continues in our own time, and we confront new 'Wild Wests', law firms and in-house law teams will need to influence new regulation, and then practice it. A lot of law is going to get written in the next few decades, and a lot of old law is going to get torn up.
But there are other more pragmatic reasons for business and in-house counsel to horizon scan and to track replacement and enhancement technologies:
- If you know how your legal advisors are using technology to expedite delivery of your work, you will be a smarter customer, you will be more able to understand the effort advisors are making to get the job done, and what a fair or market price for that is; and
- Some of these technologies may be deployable in your businesses. If your legal advisors end up using bots to get initial information needed or advance a matter, then perhaps you could use it to take questions and information from your colleagues.
Lawyers are naturally risk adverse, so new technology and ways of working are not usually within their comfort zone. To give examples from our own experience, some of the biggest difficulties in recent transactions where we have deployed technology solutions have been hampered by not being able to access the necessary client information.
Consider how your business's information is stored, is it structured in such a way that you or your panel law firm could get at it, if it was needed?
The information held by lawyers is often unstructured and to be able to usefully use the information and take advantage of technology, it will have to change.
Lawyers also like to pick holes in things; it is the nature of the work. Where a new technology is 70% effective, the tendency is to focus on the 30% that didn't work and thus lose engagement. Therefore there may be a need to keep any new developments hidden until they are sufficiently advanced and devise a communication strategy before going live.
It is also important to use the right technology at the right time. Projects that are nice to have but that don’t materially help you with delivery of your business critical projects are more likely to fizzle out, or be dropped when more pressing issues arise. If projects don’t demonstrably deliver benefit to senior leaders, they won't be 'owned' by them. Identifying projects that really do count will make a difference.
There is a risk of just focusing on the technology, its design, its intended use, while forgetting about the people who will use it. People have a range of reactions to change - including technology - ranging from denial, disbelief, anger, a wish to obstruct, through to being ardent adopters. If alongside managing the project, you aren’t also managing people and the business change necessary to embed technology and new ways of working then projects will fail.
Technology providers also appreciate the value of intelligent design in product development. The better ones recognise that if they get our insights on how we deal with key transactional processes like contract reviews, or handling data subject access requests, and if they identify what users will do intuitively, they can build better supporting technology. Increasingly, they are also recognising that we don’t want to just buy and use technology. To get most from it, we need to understand a bit about how it works, how to best build projects on it, and how to integrate it with other technology, so we are also getting technical knowledge transfer with which to upskill our own people.
Upskilling legal staff in both technology and design is now a business in its own right. At Stanford Law School they have had programme for law, science and technology for some time, but since 2016 they have also had a Legal Design Lab aimed at using technology and process design to improve fairness and efficiency in the legal system.[xvii] With the speed of change likely to increase, it is not enough to select your technology toolkit and then stop. For legal service providers whether in business or in private practice, you need to maintain the momentum.
That will require measuring the benefit of any investments made, to capture successes, but also to record the failures and to learn from them. Genuine innovation requires a lot of failure before it delivers success, and as lawyers we are not used to accepting that level of risk.
Continuing to make yourself aware of upgrades to your existing toolkit and to horizon scan for new tools is now part of the job and the candidates that you recruit to join you could also help you keep pace with developments.
It is not necessary for every lawyer to learn to code, as many of the tools are intuitive to use, but having the right skill sets or staff willing to adapt is important. This is all part of the change that is necessary to realise the full benefits of the legal technology available.
The investment necessary to embrace legal technology is therefore big and spans not just financial investment but investment of time and other resources. How long you wait before making that investment will depend on your business, but you do not want to end up in a position where:
- You can't engage in property transactions because your deeds are not digitizsed; or
- You can't win or fulfil supply contracts because your contract management system is disorganised; or
- You consistently lose litigation because the other side can access the evidence they need and you can't.
Identifying that tipping point where the investment is worth the disruption is a further decision that legal advisors are likely to contribute to in future.
Some parting thoughts which we think set out the challenges all legal advisors will face:
- Exponential change means it is likely our businesses in 10 years' time will be nothing like they are now. How do we all deal with that ambiguity and planning uncertainty?
- Expert knowledge about what the law says, how it is impacted by case law and precedent will become free and ubiquitous. So advisors won't be selling knowledge anymore, so what will they be selling?
- As we take on new technologists, people who have been through design labs or are more attuned to use technology, how do we deal with the shock of the new, how do we harness rather than resist the technological influence of the next generations?
- Lastly, in the Wild West of exponential technological progress, how do we play a part in regulation and what one futurist has termed "mission control",[xviii] how do we help?
[i] Luxton, D. (2019) 'Should Watson Be Consulted for a Second Opinion?' AMA Journal of Ethics, Volume 21, Number 2: E131-137 [Online]. Available at https://journalofethics.ama-assn.org/sites/journalofethics.ama-assn.org/files/2019-01/cscm2-1902_0.pdf (Accessed 13 March 2019).
[ii] Boggis, P and Trafford, D (2014) 'How Business Platforms Pull the Present into the Future' Formicio Limited, pp.5 [Online]. Available at http://formicio.bracedesign.co.uk/wp-content/uploads/2014/03/Formicio-Insight-Article_How-Business-Platforms-Pull-the-Present-into-the-Future.pdf (Accessed 13 March 2019).
[iii] Jirsak, J. (2017) 'Global legal services market valued at nearly $600 billion', The Global Legal Post, 22 May 2017 [Online]. Available at http://www.globallegalpost.com/big-stories/global-legal-services-market-valued-at-nearly-$600-billion-83628467/ (Accessed 13 March 2019).
[iv] Martin, T., Howe, B., Guido, J. and Joseph, F. (2017) 'Legal Technology' [Online], ACG Partners. Available at https://complexdiscovery.com/wp-content/uploads/2017/04/AGC-Legal-Tech-Apr-2017-1.pdf (Accessed 13 March 2019).
[v]Legal IT Insider (2016) 'Deloitte Insight: Over 100,000 legal roles to be automated', 16 March 2016 [Online]. Available at https://www.legaltechnology.com/latest-news/deloitte-insight-100000-legal-roles-to-be-automated/ (Accessed 13 March 2019).
[vi] Lince, T. (2018) '“No imminent AI apocalypse” - tech expert rejects predictions of mass job losses in trademark industry', Lexology, 1 February 2018 [Online]. Available at https://www.lexology.com/library/detail.aspx?g=e33f338d-1093-4f54-9792-8a75087ca1c8 (Accessed 13 March 2019).
[vii]Artificial Lawyer (2018) 'LawGeex Hits 94% Accuracy in NDA Review vs 85% for Human Lawyers' 26 February 2018 [Online]. Available at https://www.artificiallawyer.com/2018/02/26/lawgeex-hits-94-accuracy-in-nda-review-vs-85-for-human-lawyers/ (Accessed 13 March 2019).
[viii] Griffin, A. (2017) ' Robots Shut Down After They Start Talking To Each Other In Their Own Language', Independent, 31 July 2017 [Online]. Available at https://www.independent.co.uk/life-style/gadgets-and-tech/news/facebook-artificial-intelligence-ai-chatbot-new-language-research-openai-google-a7869706.html (Accessed 13 March 2019).
[ix] Kurzweil, R. (2005) The Singularity is Near: When Humans Transcend Biology. London: Gerald Duckworth & Co Ltd.
[xii] Bourns, R. (2017) 'Capturing Technological Innovation In Legal Services', The Law Society, pp.10 [Online]. Available at https://www.lawsociety.org.uk/support-services/research-trends/documents/capturing-technological-innovation-in-legal-services-report/ (Accessed 20 March 2019).
[xiii] Hanly, K. (2018) 'Free Chatbot Service Offers Aid In Contesting Parking Tickets', Digital Journal, 7 April 2018 [Online]. Available at http://www.digitaljournal.com/tech-and-science/technology/free-chatbot-service-offers-aid-in-contesting-parking-tickets/article/519292 (Accessed 13 March 2019).
[xiv] Cellan-Jones, R. (2017) 'The robot lawyers are here - and they’re winning', BBC News, 1 November 2017 [Online]. Available at https://www.bbc.co.uk/news/technology-41829534 (Accessed 13 March 2019).
[xv] IBM Cognitive Business (2017), 'How Watson Helps Lawyers Find Answers In Legal Research', Medium, 4 January 2017 [Online]. Available at https://medium.com/cognitivebusiness/how-watson-helps-lawyers-find-answers-in-legal-research-672ea028dfb8 (Accessed 13 March 2019).
[xvi] Mitchell, E. (2018) 'How Blockchain Might Transform The World Of Real Estate' [Online]. Available at https://www.blakemorgan.co.uk/news-events/blog/how-blockchain-might-transform-world-real-estate/ (Accessed 13 March 2019).
[xvii] Nagel, T. (2016) 'Stanford Center on the Legal Profession Announces New Legal Design Lab' [Online]. https://law.stanford.edu/press/stanford-center-on-the-legal-profession-announces-new-legal-design-lab/ (Accessed 13 March 2019).
[xviii] Leonhard, G. (2016) Technology vs. Humanity: The Coming Clash Between Man and Machine. Fast Future Publishing.