Insights up front
- All professions, including law, will be transformed by AI. Your competitors are using it already.
- AI is simply a way to make computers intelligent.
- AI can help lawyers do a better job in less time, by reading documents, predicting case outcomes and automating responses to routine questions.
- Any lawyer – at a small firm, large firm, or inhouse counsel — can leverage AI to drive new business models and find new sources of revenue.
- Book a free 30 minute Breakthrough Call, for independent, vendor-agnostic advice, on how to leverage AI.
It has been said that by 2030 there will be two kinds of businesses: Those that use artificial intelligence, and those that are bankrupt. This applies to law firms, and in-house legal counsel too. (McKinsey, ABA, HBR)
Between 2020 and 2025 we’re going to see truly the impact of artificial intelligence and machine learning. Many individuals from Elon Musk to Sundar the CEO of Alphabet to Mark Cuban to the CEO of IBM, have all said what I think I would paraphrase in the following, “in the next decade we’re going to see two types of companies: those that are fully utilizing AI and those that are bankrupt.” Artificial Intelligence is going to become so fundamental to the operation of a company and to its success, that it is similar to whether you use the web or email 20 years ago. Companies that are not utilizing AI in full capacity are going to miss the opportunity to truly scale and provide a product or service that is so personalized to the customer that they can’t do without it.– Peter Diamandis
Leading law firms and organizations understand this point. They know that they need to act, and the most innovative ones are excited by the possibilities.
At the same, AI is risky. By some estimates, over 80% of AI projects fail to deliver value. Why? For all its potential, it’s tricky. Unless conditions are right, it won’t work.
What complicates matters is that it is hard to find unbiased guidance about how to make AI work. The most common source of information is a vendor. As fabulous as some of the vendors are (and some are amazing) they’re biased about their solutions.
No wonder people feel doubtful about their ability to innovate. It’s hard, and they feel they’re alone. They don’t want to make the wrong choice and head down the wrong path. Plus, every hour that lawyers spend researching and worrying about what AI solution makes sense is an hour away from billable work.
We get it, and we are here to help. Tapping into our business and technical expertise, we provide vendor-agnostic advice on how to transform your legal practice and be a leader in the industry.
In this blog we will do 3 things:
- Give a short, jargon-free, introduction to AI.
- Summarize the 3 main ways that AI can improve your legal practice.
- Show you how to develop a strategy to AI that will get the highest ROI (return on investment) possible.
Part I: What is Artificial Intelligence?
What is AI? For all the complexity around the topic, the core idea is simple: AI is a technology that makes computers more intelligent. It works in three steps.
First, you begin with data. For example, in our diagram below, we begin with a picture of a cat.
Second, you feed the data (the cat picture) into the AI. The AI will take the data and to try and interpret the picture. The AI is essentially asking, “what is this?”. The truly remarkable thing about AI is that its guesses will get better over time with practice. The AI can learn! The more pictures of cats that you share with the AI, the better it gets at recognizing them.
Third, the AI makes a guess about the image and states its confidence level. In the case below, the AI is 95% confident it is a cat, and says so. After all, it might be a funny-looking dog.
In essence, that is what AI is, and how it works.
Of course, there is only so much use for spotting cats.
The real power for lawyers comes when you teach the AI to learn about parts of your job.
Part II: How is Artificial Intelligence Used in Law? (Making You Smarter and Faster)
There are hundreds of vendors in the law and AI space, everything from startups, to global companies, teaching Ais to carry out different parts of a lawyer’s job.
How can you make sense of them all, and how can you decide which ones might be a good fit for you? We suggest that you organize them into three ways that enable you to do a better job in less time:
- Reading Documents. Quickly find what you’re looking for in Discovery, Contract Review, or Legal Research.
- Litigation Analytics. Quickly assess if your client will win a dispute.
- Expert Systems: Automate the answering of common questions from colleagues and clients.
1/ Reading Documents: Legal Research, Contract Review and Discovery
A huge portion of legal work is reading documents. Usually, it is not reading for pleasure, but looking for technical information. But when we read, we don’t always catch every word. Sometimes we’re tired, or sometimes we miss key passages. Or maybe we pass on reading certain documents – you can’t read everything, right?
But an AI can. It can read millions of pages, and read them all in the exact same way every single time.
This ability helps lawyers in three main ways.
When a client problem arrives, you need to begin by checking exactly what the law is. You’ll want to find the best case that fits your facts. Legal databases now contain more than 10 million cases you can search from. This is a mixed blessing. While you’ll almost certainly have access to what you need, the problem is sorting through all the cases you don’t want to see. Doing a keyword analysis (e.g. wrongful dismissal/for executives/ in your jurisdiction) only gets you so far. It turns up the haystack when you need a needle.
So how does AI help? There are three big advantages over traditional keyword searches.
The first is by allowing what might be considered a mega-query. You can now upload your entire brief or pleadings into the document. The AI will read all of that material, and come back with suggestions on relevant cases for you. No more guessing about what combination of keywords makes the most sense. Casetext, Ross, vLex, Westlaw, LexisNexis, and Bloomberg have this feature.
Second, Casetext has another feature, parallel search. The AI can understand what you’re looking for, even if you didn’t use the precise terminology.
For example, a search about “lack of a diploma” may return a search using the term “failure to graduate.” The AI is able to understand that those mean the same thing. This feature is based upon an advanced search technology from Google.
You can even get help from AI in drafting your argument. Casetext offers a drafting tool that provides all the arguments, legal standards, and pre-packaged research you need to get things done, faster than ever. Start from a template, and have the AI prompt you on what to include, and cases to cite. This video illustrates how it works.
All of this means lawyers can do their best work, in less time. They can also stop worrying about missing some critical piece of information.
Most contract review is done by junior lawyers without the benefit of technology. Their work gets spot-checked by more senior lawyers. This is costly (think of all those billable hours) and prone to error. They can make random errors, like anyone. Unlike senior lawyers, they may not know what to be looking for.
Instead of only having junior read the contract, you can load it into a system like the one from Kira. You can ask the AI to look for over 1000 different things in the contract, such as a “change of control” clause. Importantly, as with legal research, the AI will go beyond a keyword match. There are multiple ways that such a clause could be included in the contract (something a junior lawyer might not know) and it can find these multiple versions for review. Working together, the junior lawyer and the AI are going to do a much better job than the junior alone, and in less time.
Or suppose that you are in-house counsel, and you need to know if a contract with a new vendor aligns with your company standards? You can ask BlackBoiler to automatically review the contract and give you a “Track Changes” version that highlights where it is inconsistent, and suggest edits to be consistent with your playbook. This allows you to differentiate low-risk contracts from high-risk ones. Instead of being seen as a bottleneck for internal approvals, you can approve the low-risk ones quickly, and concentrate on the more problematic ones. This video illustrates how it works.
When getting ready for a trial, it is necessary to find, understand and present electronically created information. That could include anything from an email, YouTube video or recordings captured by Alexa. There could be a massive amount of data, and searching it can be costly and ineffective. Dera Nevin, a specialist in eDiscovery, says that firms encounter the “dog problem.” If you were looking for emails about dogs, you would not find items if people didn’t use the word dog. For example, it would miss an email that said “I was supposed to take Rover for a walk but I can’t find the leash.” Or you might not even know what words you’re supposed to be looking for. Nevin notes, it is rare that people write out emails saying, “Here’s how I’m going to commit the fraud.” (AI for Lawyers, p. 100).
AI can help this process. You start by identifying documents that you know are relevant, such as some emails. You “tag” them as relevant, and the machine is able to scan the message and find other similar messages across the entire range of electronic information to retrieve those that seem most relevant. Disco and Relativity are two examples of this offering.
2/ Litigation Analytics – Getting a Forecast on Whether You’ll Win
After explaining their case, your clients are going to ask, “What is the chance that we can win?” That means you’re in the prediction business.
The prediction matters to your client. The better their odds, the more likely they are to want to litigate. The worse their odds, the more likely they are going to want to settle, without wasting all of the time and expense fighting the case.
Traditionally providing an answer meant doing research and relying on a lawyer’s experience and intuition. AI is now available that can let the data speak. It can identify which factors are going to be most important in deciding a matter and give a probabilistic forecast based on the facts in question. Suppose the case turned on whether a person was an employee or an independent contractor. The AI can review the facts and the precedents and come to a conclusion that “there is a 75% chance that the court would classify the person as an independent contractor.”
Very early in the process, this kind of information can be shared with the client, who can then make an informed decision about whether to proceed. Lawyers can also take these analyses to mediation, and use them as a means to a data-driven settlement. Bloomberg, Case-Crunch and BlueJ are examples of this offering.
3/ Expert Systems – Virtual Guides to Automate “It Depends…” Questions
Some lawyers must respond to what are seemingly simple questions on a routine basis. For example, internal counsel might be asked for a “standard” non-disclosure agreement. To a business unit, the requests seem simple – just provide whatever we normally use. Except, if there are 10 business units, operating in 20 different countries, and there are 12 different master forms for different kinds of business relationships it can get complicated quite quickly. Getting the right answer is not complicated, but it requires precision and care to make sure all the rules are followed to generate the right answer. All of this takes time and is not particularly rewarding work to do.
A better option is to teach these rules to an AI. Once it knows them, you can turn to it as your virtual guide. Think of it like a conversation with a chatbot. You start by saying you need a non-disclosure agreement, and the chatbot prompts you with the right questions – for which unit? for which kind of business relationship? for which country? With this assistance, a tedious task turns into an easy one.
There are two different ways to think about how this virtual guide can be used. One is as a tool for the lawyers to make them better at their job. They can respond more quickly to routine questions. The second option is to empower clients to use it themselves and avoid the question going to the lawyer in the first place. If the client has all the necessary information, and simply needs to be asked the right questions, relying on the AI to give guidance can free up lawyers to work on the more complicated legal work. Plus it means that clients can get access to consistent answers at any time. These solutions are provided by Neotalogic and Rainbird.
Part III: The Right Mindset to Maximize ROI from AI Investments.
The key to taking advantage of AI is to have the right mindset.
Think of it as a way to make your team of lawyers smarter and faster. With that in mind, whether you are a small firm, in-house counsel, or a large firm you can start to change your business model.
Remember, the possibilities to use AI are only increasing. Every year or two the solutions will likely double in their capabilities. That means in 5 years they could be 8 times better, and in 10 years they could be 64 times better. This is only the beginning of a rapid period of transformation.
For now, here are some steps to follow.
Large Law Firms
Start by looking for repetitive processes, and teach the AI to carry out that task. Doing so allows you to pass on a portion of those savings to your client (or reach departmental savings goals) while simultaneously increasing your margin. By automating processes, it could make sense to switch from hourly rates, to a fee for service. This would provide transparency to the client upfront about what the cost will be.
Even greater benefits will come when you start to think about the AI as a virtual version of your star employees. Teach the AI what your best employees know about contract review, or rules-based problems. You can also start to think about different AI tools can be combined for even greater effect. Imagine a chatbot asking structured questions to gather information about a case, supported by documents the AI was taught to read as an expert, and offering up a forecast on the probability of winning the case.
Virtuous circles and new lines of business can emerge. Automation leads to lower costs, lower costs lead to more customers, more customers leads to higher revenue, higher revenue funds more automation, and so on.
The introduction of AI is going to affect clients too. Their expectations will change. Today, a lawyer provides value by giving expert advice. In the future, much of that information will be more easily available, just as so much more information is available to everyone by Googling. The more easily it is available, the harder it is for a lawyer to charge a premium for it. Legal information, in other words, is going to be cheaper to get.
At the same time, as Ajay Agrawal points out in Prediction Machines: The Simple Economics of Artificial Intelligence, anything that can help an AI do its job better will increase in value (which economists call “complements”).
One thing that makes AI better is data. Firms that have lots of data that is unique will have opportunities to leverage it. That means law firms should think about the information that they have, or that someone in their network has. Perhaps it is exclusive to one industry, or one set of clients. Or perhaps it is created by merging previously isolated data sources. Either way, looking for unique data is an opportunity to find sources of value.
Another thing that makes AI better is judgment. In a world where anyone can get access to the “facts”, making sense of them is more valuable. Lawyers that have excellent relationships with their clients, where there is a high degree of trust, will be in the most advantageous position. Fostering these relationships is where the value will be created.
To be clear, not all firms are going to benefit. Firms that do not automate will be at a competitive disadvantage to those that do. While law firms are traditionally slow to adopt technology, they should be mindful of the costs if they fall behind. They might not want to change, but their competitors may force them to.
While there are not the same opportunities for small firms, like large ones, they do exist even now.
Small firms do not have the ability to hire an army of associates. This means that they are currently limited in the scope of projects they take on.
With AI research and document review capabilities, they can expand the influence of their small teams. It is like they have a virtual army of virtual associates doing the work for them. With this extra capability, are there files that you are now ready to take on?
The most innovative small firms will look ahead 3-5 years, working on the assumption that AI capability is going to continue to increase. Assuming new AI capabilities emerge, they will ask themselves, “what can we do today to be ready for these new opportunities?” Could they create a virtual star employee of their own, and could this serve a much larger range of clients, at a lower price, than today?
The volume of work is increasing as counsel deals with an increase in data, new regulations emerge, and their business becomes more complex in a global economy. There is always more to do and less time to do it. At the same time, there is pressure to increase service to internal clients, while reducing overall costs.
One way to improve the efficiency of the team is with better tools. Automatic contract review gives you the opportunity to accelerate the review process while ensuring that your playbook is consistently adhered to. With expert systems, you can create chatbots that act as virtual assistants to your counsel, or even better allow clients to serve themselves.
You should also be looking for higher levels of service from your outside counsel. You can start by asking if they are using AI, and if not, why not. You should be looking for either lower prices, or more thorough due diligence at the same price.
AI is going to change the legal profession, just like every other profession.
While some firms are certainly going to adopt the change reluctantly, the most innovative firms will see this as an opportunity.
AI can become like a star employee that takes on the rote parts of a job that lawyers don’t like to do, freeing them up for higher-level work. This will mean better service can be provided to more clients and higher margins.
If you want to discuss what it could mean for you, we’re here to help. Click here to book a free 30-minute Breakthrough Call. We guarantee that you’ll come away from it with new, vendor agnostic, insights on how AI can transform you into an industry leader.
We also suggest that you read related posts on How to Develop an AI Strategy, and AI Risks and How to Mitigate Them.