The following article stems from Dr. Gary Marchant’s keynote presentation at Womble Bond Dickinson’s Innovation Week 2021. The annual event brings together the entire transatlantic firm to discuss, consider and learn about innovation, both in the legal profession and society at large. Innovation Week 2021 is presented by the firm’s Innovation Board and took place March 22-26. Womble Bond Dickinson attorney and noted AI author Chris Mammen provided introductory remarks for the presentation.
As a researcher, academic and author, Dr. Gary Marchant is deeply involved in studying the myriad ways in which artificial intelligence impacts daily life—and will do so in the future.
But Dr. Marchant also understands how important AI is on a far more personal level. Recently, his AI-enabled smartwatch detected he was at risk of stroke before he even felt any symptoms. He was able to get medical treatment in time, and says artificial intelligence literally may have saved his life.
Such is the power of AI. Dr. Marchant said it is time to stop thinking of AI as a futuristic technology and start viewing how it already can benefit businesses, including law firms.
“It’s already here—it’s being used everywhere,” he said. But AI isn’t replacing lawyers. Artificial intelligence systems are slow learners, he said, and only are as good as the data going in them.
“It doesn’t have the human capability of common sense. It’s limited in a lot of ways,” Dr. Marchant said. “There has to be human oversight.” But while fears of AI displacing attorneys may be overblown, it is no exaggeration to say that artificial intelligence already is having a tremendous impact on how the practice of law is done.
An ABA Journal cover story in 2016 said, “Artificial intelligence is more than legal technology. It is the next great hope that will revolutionize the legal profession… What makes artificial intelligence stand out is the potential for a paradigm shift in how legal work is done.”
“This will fundamentally change our practice of law: how we train students, how we plan our careers will be affected by this technology,” Dr. Marchant said.
Substantive AI Legal Issues
Artificial intelligence already has found its way into virtually every aspect of daily life. So it should be no surprise that AI issues are becoming increasingly common sources of dispute within the legal system. Dr. Marchant identified some of the most pressing legal issues related to AI:
- Liability of machine learning systems—Dr. Marchant asks who is liable when a machine learning system’s decision goes wrong? These cases are already popping up around the country.
- Algorithm bias/discrimination—Again, a machine learning system is only as good as the data going in. So biased input can lead to discriminatory decisions in housing, employment, etc.
- Criminal risk assessment—AI is being used to determine sentencing in criminal cases based on risk assessment to the public.
- Predictive policing—Law enforcement is using AI to determine where to send patrols.
- Autonomous vehicles—Such vehicles already are on the road and as these vehicles become more common, so will legal disputes surrounding them.
- Data privacy—AI systems require massive amounts of data, which invariably will lead to concerns about privacy (facial recognition software, for example).
- Intellectual property—When a machine creates or invents something, who receives credit and ownership?
- Antitrust—When two AI collaborate without human involvement, what antitrust issues are raised?
- Medical AI—Dr. Marchant said the medical field probably is the sector that most relies on AI, and this is causing healthcare law to change as a result.
AI in Legal Practice Today
Dr. Marchant said e-discovery was the earliest legal adopter of AI, with AI systems having been employed in the field for 10 years. Numerous court decisions have supported the use of AI in document review and e-discovery. For clients, AI in e-discovery has resulted in significantly greater efficiency and cost savings.
One issue moving forward is when will courts mandate the use of AI in larger disputes. For example, a Canadian court recently capped recovery in a personal injury case, saying that the discovery expenses would have been reduced had the attorneys used AI.
“Studies have shown it’s much more cost-effective and accurate than having a human go through these massive amounts of documents,” he said. “But there’s still a human review at the end.”
Legal research is another area where AI use is common and robust. AI can review millions of documents in a matter of seconds.
“As a young associate, I remember chasing these cases and precedents and it was very inefficient. Now, you can give it to a machine to do much more cheaply and effectively,” Dr. Marchant said.
He said that unlike keyword searches on the internet, AI understands a concept and can smartly filter research documents to find relevant information, even if the exact wording may not be the same.
AI also is being used in case analytics, where machines can research and evaluate how particular judges, courts and opposing counsel operate and are likely to act. For example, how often is a judge likely to grant or deny a specific motion? How long do cases take to get to trial in a particular jurisdiction? How likely is a judge to find for patent infringement or trademark fair use?
All of these questions can be answered by artificial intelligence, giving an attorney valuable strategic insights even before stepping into the courtroom. Such information also can be incredibly useful for law firms in making billing and staffing decisions, Dr. Marchant said.
“A huge number of entities are using AI in contract analysis,” he said. “These AI systems can look through every single contract and understand what you, as an attorney, need to see.”
Dr. Marchant also said AI is better at making case outcomes than even the savviest lawyers, using data to make estimates and calculate odds.
“There are even analytics that tell you how to frame a brief for a particular judge,” he said.
AI in the Courtroom
AI in the legal arena isn’t the sole domain of law firms. While courts typically are thought of as being well behind the technology curve, some have successfully employed AI for a variety of functions. For example:
- Automatic scanning and docketing of court filings;
- Using Facial recognition log-ins for court computer systems;
- Employing chatbots for public inquiries. In the near future, legal chatbots seem almost certain to become a commonly used technology. Such AI-powered features interact with visitors to a courthouse or law firm website to interact with clients in real time.
- Administering automated, AI-enabled juror hotlines; and
- Identifying “red flags” in guardianship and conservatorship cases.
And just as attorneys can use AI to sort through massive amounts of information to pull out critical data, judges already are employing AI to help manage data-heavy caseloads. Human judges still make the decisions, but trained AI review documents to gather and synthesize the information that will be useful to the judge in making those decisions.
The Chinese legal system has taken it one step further, using holographic AI judges to decide actual cases from start to finish. While it may sound like something out of a science fiction novel, “this hologram-judge sets schedules, asks questions, takes evidence, and issues dispositive rulings,” according to a December 2019 New York Law Journal article. Other nations are considering AI judges as well.
One area of concern for the courts, Dr. Marchant said, is that AI can create amazingly convincing forgeries of videos, photos and documents. He said it will become increasingly difficult to determine whether courtroom evidence is legitimate or fabricated.
AI in the Law: Challenges and Upsides
For law firms, Dr. Marchant said that early adopters of AI could reap the rewards that come with offering a better, more efficient product. For example, a corporate client may choose to send all of its contract review business to a firm that employs AI review.
“There’s sort of this law that if can do things more efficiently, you’re going to get asked to do more things,” he said. “It allows the profession to do more things for more clients.”
Similarly, if firms are able to reduce costs to clients by using AI, clients that may have been on the fence about engaging them due to price may be able to afford more legal services.
Dr. Marchant also noted that AI and related technologies are transforming virtually every sector of business. Those companies will need—and expect—lawyers well versed in AI and the law to counsel them in the years to come.
“If you get behind, you’re always going to be behind. The early adopters are going to get a lead and it’s going to be hard to compete with firms that have this technology.” he said.
But there are challenges. To date, AI vendors focus on specific legal tasks. There are no across-the-board legal AI systems (at least, not yet). So law firms may need or want to work with many different vendors to provide various AI services.
The nature of machine learning means it takes time for AI systems to learn a firm’s data and operations. So it would be highly disadvantageous for a firm if its AI vendor went out of business, Dr. Marchant said.
There also issues of malpractice to consider—both related to relying on AI and to not relying on AI. How do we know when we should rely on AI? That is a difficult and always evolving question to answer, Dr. Marchant said.
In addition, the question of data ownership remains murky, Dr. Marchant said. Who owns the data used to train an AI—the firm, the attorney or the vendor? When an attorney changes firms, what can they take with them, particularly when training attorneys on AI use is such an expensive proposition.
Law firms also will have to reevaluate how new associates are utilized, given that AI is taking on the roles of document review and legal research those lawyers traditionally have done. The value of more senior associates remains strong, but how first- and second-year associates fit into the new model remains to be seen.
One thing is for sure, however: AI is here—and it is here to stay.
“It’s the combination of the human and the machine together that is the most powerful tool,” he said. “One thing is becoming clear: lawyers who use artificial intelligence will replace lawyers who don’t.”
For more on this topic, click here to read “Artificial Intelligence and Patent Law: What Happens After DABUS?” co-authored by Womble Bond Dickinson’s Chris Mammen and Brent Babcock.
Also, click here to read “Womble Bond Dickinson Innovation Week 2021: How the Next Decade’s Technological Advances Will Change Business,” based on a presentation by Wired Magazine Executive Editor Jeremy White.