“I assume you are all here because you want to find out how long it will be before robots take away your jobs?” attorney Bob Ambrogi jokingly asked the crowd at the start of the Artificial Intelligence (AI): Why It Matters to Your Future Bankruptcy Practice session at the 2018 ABI Annual Spring Meeting on April 20 in Washington, D.C.

As AI continues to expand in consumer technology (such as Apple’s Siri or Amazon’s Alexa), financial services and even professional sports, Ambrogi led a panel of expert lawyers to discuss what AI is and to cut through some of the myths, fears and hype surrounding AI in the legal industry.

“The tools that all of us build are designed to augment and accelerate the work of lawyers, rather than to replace the work of lawyers,” said panelist Owen Byrd, chief evangelist and general counsel of Menlo Park, CA-based Lex Machina.

Karim Guirguis, Chief Strategy and Innovation Officer, American Bankruptcy Institute
Karim Guirguis, Chief Strategy and Innovation Officer, American Bankruptcy Institute

AI Is a How, Not a What

“It’s just software,” Byrd said. “Powerful, robust software.”

He highlighted machine learning and natural language processing as two components of AI software which will be effective and useful tools for the legal industry. The software involved in machine learning enables a program to comb through large data sets, such as dockets, and to learn words and phrases to add to its store of knowledge, according to Byrd. This enables it to complete subsequent searches with better precision.

“We’re not talking about the boy robot who learned to love,” Darby Green, commercial director for Litigation and Bankruptcy at Bloomberg Law, said of machine learning. “Think about big pieces of data, like court opinions and dockets. What the machine-learning algorithms can do is recognize where patterns occur over and over again and start to draw conclusions and improve.”

The other key to AI technology in law is natural language processing. “It’s just like it sounds,” Byrd said. “It is a way of using software to process natural, verbose, unstructured language.”

Even so, some legal professionals might have the wrong idea of what AI can mean for their practice, according to Michael Mills, co-founder and chief strategy officer of New York-based Neota. Upon encountering lawyers who think, “We ought to get some AI in our law firm,” Mills explains that AI is not a thing law firms can buy. “AI is, in fact, a how,” Mills said. “It is a way of thinking about and a way of building software to solve a set of problems.”

John Hartgen, Public Affairs Manager, American Bankruptcy Institute
John Hartgen, Public Affairs Manager, American Bankruptcy Institute

A Connection Is Made 

According to Green, AI software is making the legal industry more efficient by taking on rote tasks, thereby freeing lawyers to focus on more strategic assignments and business development. “It’s making connections between the data to make it more meaningful,” she said.

Ed Walters, CEO of Washington, D.C.-based Fastcase, said regular lawyers, not computer engineers, can now use artificial intelligence as a software tool to accomplish increasingly ambitious goals. In litigation cases, AI offers lawyers a chance to step away from anecdotal-based answers to ones that are based on data clients can use to answer difficult questions such as, “Should I accept this settlement offer?” or “How much risk do I face?”

Clients already make many decisions such as marketing, hiring and budgeting based on hard data. Walters believes the next generation of lawyers will be delivering data-based legal services. “The answers aren’t anecdotal; they should be data-based answers,” he said. “We haven’t had access to those answers in the past, but in our near future, because of these technologies, we will have answers to those questions that are based in data.”

While big firms were primarily the first to invest in and use AI systems, Thomas Hamilton, VP, Strategy and Operations at San Francisco-based Ross Intelligence, finds AI to be more of a “leveler” for small law firms. “I don’t think that big law hiring numbers will decrease because of AI,” said Hamilton, but smaller firms armed with AI will be enabled to take on new and different types of work.

Byrd agreed AI software helps level the playing field for legal services. “Now you can get high-quality, reliable information from any of these different providers that gives you the same level of horsepower that [big law firms] have,” Byrd said. “I think that’s actually an advantage to boutique firms and smaller practitioners.”

Mills advised practitioners to look at the ways their firm generates revenue and profit, and to think about the ways in which they see themselves applying AI technologies to solve specific problems. “The partner who walks in and says, ‘We need some AI,’ is asking the wrong question,” Mills said. “The partner who says, ‘I have a practice in which we have the following challenges from our clients and from our competitors: how can these tools help?’ is asking the right question.”

More Data, Fewer Hunches

Panelists also pointed out the ethical requirements AI may soon usher into the legal industry as access to the technology spreads to clients and courts. “We’re not far off from an ethical obligation to check the numbers before you litigate in front of a judge because your in-house client is,” Byrd said. “They are supervising your work, and they’ve got access to these tools.”

“In the past, when everybody was using hunches, if you used a hunch and got it wrong, nobody would second-guess you,” Walters said. “If your clients are using data to figure out whether a settlement is a good idea or not, and you’re using hunches, and you make a bad decision, they are going to call you out on it.”

Mills said practitioners must be aware that clients — and the courts — will have access to AI systems. “I know that when I was a law clerk to a federal district judge, the first thing the judge did was ask me to check this brief to see whether it had been wisely and prudently and fairly populated with citations,” he said. “Now that process can be automated very quickly.”

Green pointed out a few AI-related ethical considerations lawyers should be thinking about, including duty of competence, duty to supervise and the unauthorized practice of law. She said 31 states maintain a duty of competence, with an affirmative duty for lawyers to both understand the technology available and to consider using the technology if it’s going to support clients. “AI and all these other technologies are not going away, and the bar is recognizing that,” Green said.

One of the most frequent questions Mills hears while talking to general counsels of law firms is how to establish who is responsible when AI misbehaves. Before he explains how his firm tests and measures AI, he asks about the processes, procedures and systems in place in the prospective firm to test the quality of work the lawyers do across the entire firm. “We can give you statistically defensible answers to the question about how good our software is, but you’re asking that question in the midst of a profession that doesn’t do a really good job of measuring and evaluating and dealing with the consequences of less-than-perfect conduct.”

To Infinity & Beyond

Comparing the shift toward AI in the legal industry to Ford’s adoption of the assembly line during the Industrial Revolution, Walters envisions an expansion of legal services made available to more Americans. “If we change the way we work, if automation, if specialization, standard parts and, yes, artificial intelligence help us to work smarter, just like Ford did [with the assembly line], then we can reach more of that latent market,” Walters said. “We can make the market of legal services even bigger. We can provide better, even more intelligent services.”

Paraphrasing a quote from Steve Jobs, in which Jobs likened the personal computer to a bicycle for the mind, Hamilton said AI is much like a bicycle for the lawyer’s mind. “We direct the bike, we power the bike, but the bike allows us to do more than was ever before possible.”

Walters concluded the session by offering practitioners a way to view AI going forward. “Give to the robots the parts of the practice that are most robotic, and let us keep for ourselves the parts of the practice that are most human,” he said. “That’s the real potential for AI.”