“Alexa, Bill My Time:” How Artificial Intelligence Makes Lawyers More Efficient and Reduces Costs

“Alexa, Bill My Time:” How Artificial Intelligence Makes Lawyers More Efficient and Reduces Costs

5.15.2020

By Brandon Vogel

The robots are now in your law firm.

Well, not exactly, but artificial intelligence (AI) is the next wave of software and it could make your life simpler. Technology that we saw in “The Terminator” is now in play and helping lawyers reduce their legal research costs by up to 80 percent.

Thomas Hamilton, vice president of strategy and operations at ROSS Intelligence, said that “artificial intelligence has hit that sweet spot where lawyers can do more quickly for their firm, than ever before” on the recent CLE webinar, Making Legal Research Profitable and Effective for Your Firm During COVID-19.

What we thought only humans can do

Hamilton started off at Dentons, now the world’s largest firm. He saw firsthand how technology transformed its business operations. He was surrounded by companies that focused on natural language processing to increase efficiency and accuracy in legal research. His “aha moment” came when he realized that artificial intelligence software “can do something that we thought only humans could do in the past.” He left the practice of law to become a legal technologist.

He said his current company ROSS is a legal research platform powered by advanced natural language processing/machine learning models. It augments a litigator’s skills and reasoning with curated collections of the most relevant case law and legislation. Its core mission is to help everyone obtain the best possible legal outcomes by keeping litigators up to date with case law and simplifying research.

He explained that most existing AI applied to natural language processing, which, in its simplest terms, is how humans and computers interact with each other, particularly how to program computers to process and analyze large amounts of natural language data.  “We are now able to communicate with computers the way we do people,” said Hamilton. “That makes it a very exciting development for the law.”

Natural language processing applies to due diligence, contract review, discovery, litigation analytics, document automation and billing.  At a high level, AI involves spam filters, voice to text features, smart personal assistants like Alexa and Siri, and fraud detection software.

He acknowledged that legal research costs are typically high. “The key problem with legal research is that ends up pricing a lot of people who could otherwise purchase legal services” said Hamilton. “Less demand for legal services makes it harder to hire a lawyer. Eighty percent of people who need legal representation cannot afford a lawyer. A lot of that ties into inefficiencies into how legal services is delivered.”

As a practicing attorney, most of his time researching could not be billed to clients. He noted that clients have become much more cost-sensitive after the 2008 recession.

He explained the shift in how technology has affected dentistry and the patient experience. “When you go to see the dentist now, you spend very little time with the dentist. At some point, the dentist comes in to check in with you, but you spend most of your time with the hygienist,” said Hamilton. “It allows more affordability and lets the dentist focus on what they like to do”.

Hamilton said that as law continues to expand so does the length of documents. As an example, the 2003 Supreme Court case of McConnell v. Fed Election Commission was 50,000 words long. “As this trend continues, it becomes harder to find your answers without software,” said Hamilton, who noted other issues include a lack of headings, unclear language, dense prose or no unified citation standards.

Time is money

Andrew, a solo attorney, has limited time and resources and past cases. He largely works on contingency. For his typical case workflow, he accepts the initial assignment, performs his primary research and then decides if it’s worth taking. The research can take up to 50-80 percent of his time.

“The issue is that we all know time is money. Your clients can’t always pay you based on that time you are researching,” said Hamilton. “This means you have to pick and choose clients.”

Google searching has its limits. “There is a lot of complexity in terms of the user correctly searching using Boolean searches with quotations and clarifiers like AND and OR,” said Hamilton. “Traditional legal research tools also put the entire onus on the user to find the best cases. No responsibility is taken.”

ROSS has a created a system where you ask questions as if you were talking to a human. It allows for information extraction, summarizing and named entity recognition. Legal research costs can be reduced significantly.

Existing search systems do not understand natural language, per Hamilton, so they return mountains of results, which lawyers waste hours going through. Documents are retrieved by matching keywords and ranked by frequency.

ROSS predicts the likelihood of an answer by including synonyms, semantics and dependencies in addition to key words. It will help you find similar language, fact matching and analyze documents.

For example, in the sentence, “The red dog ate the cat,” AI can understand the dog is specifically red.

In ROSS, you can pose the question: What is the standard for false necessary implication advertising in New York since 2010? It will automatically filter out cases before 2010 and include entire passages in the summary.

“It’s not just keyword matching but looking at synonyms,” said Hamilton. “Through fact matching, you can narrow it down to a single case that answers your question.” In this instance, it pulled up Playtex Products LLC v. Munchkin, a 2016 federal case involving false advertising claims.

“It reduces the legal research spiral,” said Hamilton. “It takes you there more directly in a straight line. “