How AI and Machine Learning Enhance the Legal System
AI technology is lowering costs while increasing efficiency and productivity in legal practices across the world.
Artificial intelligence (AI) looks to be the most disruptive class of technologies in driving digital business forward during the next decade. Yet, even amongst the most tech-savvy professionals, there is divided-opinion over what AI can or cannot be defined as, and what it can and can’t achieve.
The most exciting aspect of applying AI in the legal profession lies in the automation of repetitive and straightforward tasks, like eDiscovery or judicial bill review, while enabling human experts to improve results beyond what machines or people could do alone.
This combination allows for enhanced productivity while driving significant time and resource savings. This piece will clarify what AI is, how it’s being used today, and how it can improve legal operations — now and in the future.
The goal is to help you understand the broader context in which AI-enriched legal software is being developed and how you can benefit.
Artificial Intelligence Is at Peak Hype
Research and advisory company Gartner published a Hype Cycle, which tracks emerging information technologies as they move toward mainstream adoption. It helps companies distinguish hype from mainstream adoption, and helps to set an expectation for when this technology will be in widespread use.
Artificial intelligence was one of three huge trends in the 2017 Hype Cycle. Gartner called this class of AI technologies “AI Everywhere.” It emerges out of the “radical computational power, near-endless amounts of data, and unprecedented advances in deep neural networks” that are affecting technology development across industries.
Unsurprisingly, most of the technologies are already at the “Peak of their Expectations,” meaning that they are talked about as the “next big thing” but have not yet been adopted by most consumers. It is important to note that hype isn’t a bad thing.
To say that there is a lot of hype around a given innovation is not to say that it lacks value. It simply means that potential consumers of these technologies should be careful. When considering a technology that’s currently at the Peak of Inflated Expectations, it is wise to vet the solution carefully, making sure that it does help you meet your organizational goals.
Ask the vendor to show you evidence of what their product can do. Be confident they can back up claims about the hyped technology with practical, real-world facts.
No solution should utilize AI without good business and technology reasons. Because of the hype currently surrounding AI, many vendors are marketing it without providing a clear and defined description of how it benefits the end-user. In a later section, we’ll discuss vendor evaluation criteria to help understand if a particular AI product will help you.
Understanding What Artificial Intelligence Is There is a good deal of confusion about AI across industries. These points will help to clarify some of that confusion:
There is no single, ‘finite’ definition of AI; Wikipedia describes AI as:
“intelligence exhibited by machines, rather than humans or other animals.”
Artificial intelligence is a collection of technologies — not one single technology. For example:
Machine Learning exposes computer programs to large data sets so they can find patterns. Feedback and repeated iterations allow the programs to learn, improve, and deliver increased accuracy.
Natural Language Processing (NLP) is a way for computers to break down written or spoken language into concepts and entities and to build relationships to analyze language like any other data type.
Robotic Process Automation provides computerized workflow and applications that can take over the mundane, repetitive — and often boring — aspects of various business processes, so they require limited attention from people.
The specific mix of technologies employed in solutions is dependent on whether they are the right tools to help solve the particular problem at hand.
The Role of AI in Legal Operations
All Artificial Intelligence is Not Equal
As Jerry Kaplan noted at MIT Technology Review, AI has a PR problem. Despite breathless reporting, he says, accomplishments written about in the mass media are not evidence of significant improvements in the field, but are stories ‘cobbled together from a grab bag of disparate tools and techniques’, some of which may be considered AI, some not.
The scope of AI, as widely reported in the news, ranges from cyber-surveillance and facial recognition to assistive robotics, and toothbrushes.
Where Does Machine Learning Fit In?
Machine learning is described as the field of computer science concerned with giving computers the ability to learn without being explicitly programmed. Machine learning is an algorithm or a statistical model that learns to:
• Recognize patterns in existing data, then;
• Predict similar patterns in new data. The more data you have available, the “smarter” the program gets.
The program essentially “trains” itself by repeatedly combing data and learning from the patterns it finds. The next step after identifying patterns in existing data is the ability to make predictions.
As an example: A restaurant could, for example, better cater to their customers by building a machine learning model that analyzes their busiest periods, the most popular menu items, and estimated waiting times to more accurately schedule service staff and schedule stock supplies for improved customer experience.
As the business expands, and customer needs evolve, the model can adapt to the new situation and make further recommendations, all based on brand new data.
With continuous machine learning comes constant adaptation and improvement. When machine learning is implemented for legal operations, the data from invoices, matters, other legal records “trains” the AI to recognize patterns, while the expertise of the legal department staff provides feedback that allows the AI to improve results over time. With these things in mind, let’s now consider how intelligent today’s artificial intelligence is.
How Intelligent is AI?
The idea that machines can actually ‘think’ is the central tenet of AI in popular culture, including movies. And robot narratives rarely turn out well in movies. Kaplan notes,
“Had artificial intelligence been named something less spooky, we’d probably worry about it less.”
He continues to say that while it’s true that today’s machines can credibly perform many tasks (playing chess, playing Go, driving cars) once reserved for humans, that does not mean that machines are growing more intelligent or ambitious.
It just means they’re doing the things we built them to do. Essentially, AI programs are one-trick ponies that specialize in and excel at, one single task.
The Path from R&D to Market Value
Breakthroughs in hardware, software, and techniques such as deep learning (where a machine gains abilities from experience) came in 2013 and 2014. Thanks to newly available computational power, massive data, and better algorithms, AI made significant jumps forward as a field.
Machines could now recognize objects and translate speech in real-time. Investments that companies had made some years earlier began to pay off.
The Role of Artificial Intelligence in the Legal Industry
During 2014 and 2015, a steady stream of PR revealed AI as the secret recipe behind Amazon and Netflix recommendations, and Facebook’s image recognition. Virtual assistants became increasingly popular such as Apple’s Siri, Microsoft’s Cortana, and Amazon’s Alexa, Google’s smarter search results, and more.
As the coverage grew, AI made headline news — both positive and negative, with prominent scientists and technologists calling for research into the potential societal impacts of AI. And then mentions of AI surged in company earnings calls in 2015 and 2016 as business leaders rushed to acknowledge the importance of this burgeoning technology.
This helped to begin putting a real value on AI technology. Legal departments are now seeing the potential in AI, and some have started to adopt it, but this is happening slowly, as discussed by an expert panel at the CLOC Institute.
Due to the fact that the term “artificial intelligence” typically is not well understood on a technical level outside of the software industry, some legal professionals have a vague notion of it as future technology that isn’t yet ready for use in a legal environment.
Or they may consider it as a general term for using computing power and data in a way that blurs the boundary between what machines and humans can achieve.
And even as lawyers develop a better understanding of AI’s practical uses, they aren’t always eager to adopt new, unfamiliar technology.
While some lawyers are concerned about their job security in a legal department where AI is used, there is no need to worry.
According to the CLOC Institute panel members, to be effective, AI requires people, processes, and technology working together. In a legal environment, this means:
• Experts that define solutions and help to oversee its implementation
• A service model that can deliver work in a way that meets the client’s and departments’ needs
• AI technology that can provide information tracking, reports, and workflow to improve efficiency
AI in the Legal Department
For many legal departments, the first instance of AI used is in the context of eDiscovery. The huge amount of data that must be reviewed during any eDiscovery process makes it difficult for lawyers to adhere to aggressive schedules.
Also, few attorneys want to spend weeks running searches and sifting through results if they know they can count on automated technology to correctly complete the task for them.
Today, more sophisticated machine learning technology has become common in popular eDiscovery solutions. The continuous success of AI technologies in the area of eDiscovery has proven that AI can be used as a valuable tool in law. It is also massively improving over time as algorithms are refined and AI applications “learn” from their previous work.
Using AI the Right Way
Some companies are so eager to take advantage of the AI and machine learning “buzz” that they may leverage the technology in ways that are not helpful to clients.
This often leads vendors to emphasize the specific algorithms and analytics that their solution employs, rather than focusing on the benefits and outcomes.
Gartner recommends a few questions to ask when talking to that type of vendor to help you to determine whether an AI-based solution will genuinely help you meet your goals:
• What aspects of machine learning does the solution use? Ask them to explain specifically how machine learning is utilized and how it improves ROI.
• What analytics methods does the solution use other than machine learning? They should be able to describe how any other analytics complements the AI and the value each delivers.
• Can you use our data to provide a Proof of Concept? This is the most direct way for the vendor to demonstrate exactly how their solution can improve your results.
The ultimate goal is to understand both how the inclusion of AI makes the product superior to non-AI options, and precisely what added benefits you can expect if you implement it.
Through utilizing machine learning and AI, legal and claims departments see that compliance with their billing guidelines — an issue many companies struggle with — can be improved significantly across outside counsel relationships. Billing guidelines are an essential component of ensuring law firms adhere to your requirements on staffing, process, and legal practice, as well as invoicing.
While AI offers excellent opportunities in cost, efficiency, and productivity benefits, it is our experts who bring deep insight into our customers’ most challenging problems. This expertise, together with the evolving artificial intelligence technologies we’ve covered in this white paper, promises to continue revolutionizing and empowering legal operations