Artificial Lawyer Is 1 Year Old: The New Era of Legal AI

Artificial Lawyer is celebrating its First Birthday and what an amazing 12 months it has been covering the world of legal AI and advanced automation.

One year old this month

So much has happened, so much legal tech innovation has taken place, so many inspiring people have come forward as AI pioneers. And because of this the legal world is a little different today, though far greater changes are to come.

Sometimes a year in the legal market can pass by almost unnoticed because so little that is significant happens in terms of market change. And then there is the other type of year, when sometimes each month feels like an historical epoch because so much happens. That was the kind of year Artificial Lawyer and many in the legal AI world just experienced.

Ideas about the potential market size and use cases of AI in the legal sector, which seemed experimental and relatively narrow in early 2016, now just a year later seem quite practical and also much more diverse. What started out very much as a linkage between doc review and legal AI has diversified into an array of applications that include (among others) –

  1. Due diligence/lease analysis/employment contract review/compliance projects.
  2. Litigation doc review/eDiscovery/privilege review.
  3. Legal research/internal data core management/KM search.
  4. Litigation/court decision outcome prediction/patent strength testing.
  5. Triage/intelligent interfaces and expert systems
  6. …(and about at least five other significant areas of AI application in the legal world at last count)
When legal text is data…

As Artificial Lawyer has remarked several times over the course of this year on seeing pioneers solve new problems using AI systems, the limits of AI in the legal world are mainly set by our imaginations, not the technology.

In a world where legal text is effectively just digital data that can be processed at lightening speed, extracted, sorted and applied to a myriad of use cases, then we are now living in a different world. In fact, it may not be hyperbolic to say that the legal world has crossed into a new era of true digitisation, with natural language processing (NLP) and machine learning turning a world of unreadable data (i.e. legal documents and communications) into a gigantic data set that will set free and inspire dozens of other AI applications in the years ahead.

Meanwhile, AI companies that seemed to be on the fringes of the legal market, even on the fringes of the legal tech market, are now front and centre. Larger legal tech companies are merging with smaller legal AI companies to seek powerful synergies (special shout out to iManage and RAVN, obvs…), the number of conferences dedicated to, or specially featuring, legal AI, has gone from one or two a year, to what must be well over a dozen or more. All in 12 months….

Just the fact that Artificial Lawyer now has a global readership and a community of people from all over the world who regularly send in press releases and emails to share the news of the birth of a great new legal AI company, or a new application or a new iteration of an existing AI system, is a sign of the times.

And, it would be great to list here all the brilliant legal AI and automation companies out there that have been covered in the last 12 months. But, to be fair to everyone it would take up most of the rest of this special birthday issue, which in itself is a sign of the times. We are looking at dozens of great companies here in this space.

The people part of all this has also been fundamental. Technology doesn’t create itself, nor promote itself to clients without some human intervention (not yet anyway…). Artificial Lawyer has had the privilege to get to know a large number of inspiring people working in this field, from founders with a creative gleam in their eyes and mind-boggling AI tech experts, to the many people that bring attention to this new and growing industry such as other writers on the subject (special shout out to Fredrik Svärd!) to conference organisers, to the funders, to the students and academics, and of course the savvy lawyers, innovation directors, legal engineers and IT directors who are using this new technology.

What an awesome community of people! One of the best things about the artificial intelligence world is the humans out there that are making it happen…ironic, huh? Thank you to everyone in this industry. It’s been a pleasure getting to know you.

OK, so, Artificial Lawyer is One Year old, the legal AI market is growing nicely, everything is going at full tilt. Now what? Well, I thought it would be a good opportunity to share a couple of observations on where we have got to in the middle of 2017 and what it means to be where we are now.

But before we do that, please note that Artificial Lawyer is taking a few days off this week to recharge his batteries. And if the fairly long article today leaves you hungry for more while I take a short break this week, then please check out the two video interviews I did with: Nick Rishwain + Ivan Raiklin at LegalTech Live  in the US (this June) and also Stevie Ghiassi at Legaler in Australia (done in March, but just published now).

LegalTech Live interview (June 2017)

And,

LegalMeets, Legaler (published June, recorded March 2017.

But, back to the key observations about the legal AI world……

The New Era of Legal AI in 2017

Catching Up With Change

What came across very clearly from all the legal tech events Artificial Lawyer has attended/spoken at recently was that there is now a sense of needing to catch up with where legal AI has got to, rather than wondering when it would become relevant to firms and clients.

One might say that as some firms have waited to see where things are going they’re now waking up to the realisation that many of their peers, which they had assumed were also just watching things evolve, have in fact started piloting legal AI systems.

When you get a chance to chat to conference attendees in the breakout sessions and afterward at the drinks, that’s where you hear all the stories. And the message is quickly apparent: there is a lot more activity in the background going on than is reported in the press.

In part this is because, quite wisely, many firms don’t promote the fact they are doing pilots, or doing their first few pieces of client work with AI. Press releases may come later, but lawyers at other firms would be naive to assume that because a rival isn’t featuring in the press, yet, that they are not already experimenting with AI, or now starting to do client work with it. After all, clients are now asking law firms in RFPs to set out if they use AI systems and how will this help the corporate client. That’s got to have an impact across the market.

Hence, this new sensation of ‘catching up’ among those some law firms that thought that perhaps they still had two or three years to get their heads around this before needing to take it seriously. Time’s up.

Big Law, Small Law, All Law

Legal AI is often associated with ‘Big Law’, but again, go out into the conference breakout sessions and talk to people and you hear an increasing number of firms of relatively small size, e.g. well below £50m ($63.5m) revenue, are also piloting legal AI systems.

There are examples of law firms as small as two partners making use of AI systems. And whenever Artificial Lawyer asks an audience of smaller law firms if they see AI as relevant to them and their clients right now, the vast majority say: yes. For the larger firms, they don’t even need to be asked in most cases.

The other factor is how global the legal AI world is now. Naturally there is a lot going on in key legal AI centres such as London, Toronto and San Francisco. But also across the US, such as in New York, Philadelphia and Chicago. Then across Europe, where Germany, France, Belgium, Holland, Sweden and Finland are all charging ahead. Then we have other centres with fewer legal AI companies, but which where there is a surge of interest such as Israel and Singapore. And of course, we have Australia and New Zealand. Mainland China is now also beginning to explore legal AI – and the outcome there will be fascinating to see given the nation’s economic importance. And, last but not least, legal AI companies have begun to evolve in India too. This really is a global phenomenon.

Raising the Debate

Artificial Lawyer on the conference trail, chairing a legal AI panel at CogX

What is perhaps most encouraging is the fact that whereas as a few years ago the debate was often about ‘Will I lose my job as a lawyer to some sort of robot?’ the bulk of the questions at meetings with law firms and at conferences now are practical in nature.

Some firms are still stuck on the existential crisis aspect of AI, but even these partnerships soon climb down from the ledge and start to get excited about legal AI once they know more about it. Rather than worrying about AI, they soon start to ask very practical questions about how to use it. Questions such as:

  • How can I integrate doc review AI systems with the rest of the software my firm uses, or other new types of automation system, to create greater value? (I have a few ideas on this….)
  • How will AI impact the training of junior lawyers? (The answer to this one is: if you want doc review or very basic legal research to remain the key source of legal training for your young lawyers so they can go out and advise clients, then your firm’s got problems….)
  • What practice areas can we use AI systems in? (The answer: there are likely very few areas of commercial legal practice where some aspect of AI cannot be of use. First you need to consult internally and find out what is best for you. Also got some more ideas on this one…)
  • How do we get the partnership to make use of AI across the firm? (Great question. Building a team that can act in a quasi-consultative way and broaden knowledge across the firm is part of the solution.)
  • Is AI expensive…..it looks like it will be? (The answer has to be that associates on high salaries doing low value work and rooms full of paralegals that you don’t really need are expensive. Moreover, some AI doc review companies are working on a pay-what-you-use basis, rather than charging very large upfront costs. There will also likely be price competition as more firms and clients take up AI doc review in the months ahead.)

Keeping Positive

Amid all social or sector change there are twin currents of discourse. One is positive and optimistic, looking to the benefits, rejoicing in the positives, marvelling at the new. The other is negative and cynical, looking for faults, seeking out where there may be problems that can be exploited for gain.

Hmm…something seems familiar

Personally, I prefer the former line of conversation when it comes to AI. But, it’s natural that some don’t. Moreover, some people believe there is a pile of cash to be made in looking for problems in AI and then selling ‘advice’ on it to large corporates that scare easily. It all seems very ‘Y2K’…remember that old money-spinner..? But, we just have to power onwards with the positive while addressing the issues as they come along.

We could get into a great debate here on this topic, especially on the misunderstanding about the difference between algorithms created ‘by hand’ that seem to fulfil a need to pass the buck on making decisions, and using AI systems to analyse huge amounts of unstructured data to help build accurate and impartial models that reflect reality and going with their outcomes because, well….they reflect reality. I.e. one is snake oil, the other is the product of science.

To the untrained eye, medicine and snake oil both appear to be the same thing, but clearly they could not be more different in terms of their results. Moreover, don’t assume that just because an organisation has created some sort of algorithm to come to an answer on a problem that there is any AI analysis of a large dataset in there. Anyone can make a formula on the back of an envelope without any real data analysis that says ‘If person X has Y attribute then Z is the outcome our organisation will choose’. To some degree many companies do this already, just look at some corporate hiring policies that automatically reject certain candidates without even meeting them, for example, though those decisions may not have any scientific or proven rational basis. Saying that it’s the ‘algorithm’ creating the bias inside an organisation is often to get the wrong end of the stick.

But, Artificial Lawyer is confident that this speed bump of misunderstanding will be overcome. That said, as an industry, legal AI and AI in general, cannot avoid the debate. We’ll need to meet it head on.

Conclusion

As you can see, what started out as an unknown land where strange beasts called ‘legal AI start-ups’ lived has now become an official part of the legal mainland, if not in fact at the very centre of it.

Legal AI has been so successful already because the bigger story is always one about positive change, the one about using technology to solve problems, deliver value, open up new areas of insight and capability, as well as to boost productivity, not just in the law, but across the entire economy.

Thanks to everyone Artificial Lawyer has interacted with over the last 12 months. It’s been a real pleasure. And the best thing of all…….this is just the beginning of a new era of legal AI.

Thanks very much for reading. I’m off now for a week and will resume publishing 3rd July. The adventure continues…