A good lawyer knows the law. A great lawyer knows the judge.
This old saying is true. More times than one could count, local lawyers from small firms have decimated high-priced attorneys. Why? The local lawyers have appeared before the judge 100 times. They know the evidence that will be accepted and rejected by the judge. They are familiar with the preferences of judges when it comes down to expert witnesses. BigLaw lawyers in major cities are no different. They are members of the same country clubs that the judges. They regularly appear before them. Every outcome is documented.
Local and/or BigLaw attorneys have always had an advantage because of their knowledge.
Until now.
Lex Machina is the great equalizer.
Lex Machina allows users to enter the name and case number of a judge, which will show them all their previous rulings on similar cases. Lex Machina, for example, can show attorneys the rulings that the judge who will soon be presiding over the case has made in similar cases.
The software can tell you which law firms were most successful before that judge in similar cases, what type of evidence the judge accepted or rejected, and which expert witnesses were most successful. The push of a single button can complete thousands of hours of research.
Bob Roberts their CEO was asked How this all is done.
Lex Machina provides insights on the judge, court, opposition counsel and similar cases in the past to help lawyers and corporate counsel develop their trial strategies. The platform analyzes data from millions federal and state courtcases, including information about judges, parties and law firms, as well as outcomes.
Lex Machina allows lawyers and corporate counsel to better understand how judges have ruled on similar cases. It can be used to develop a better trial strategy, by customizing their arguments and evidence according to the preferences of the judge.
Lex Machina also provides information about opposing counsel, their litigation history and allows lawyers and corporate counsels to anticipate and respond more efficiently. The platform allows users to compare the success rates between different law firms and lawyers, giving them insights into those who are most successful in certain areas of law.
Mr. Roberts continues.
“Lex Machina provides invaluable outcome analytics that provide data-driven insight on timing, findings and resolutions, as well as damages. This helps attorneys predict how long a lawsuit could take, and what the likely outcome is. This information is useful both for crafting litigation strategies as well as accurately advising clients and managing expectations.
The information will help attorneys understand the opposing counsel and judge, and also determine if the case should be settled or if the in-house lawyers at $500 per hour are better suited to the case.
Bob said that AI isn’t just for judges and lawyers. Even expert witnesses can be selected with the AI!
The Expert Witness Explorer is one of Lex Machina’s Quick Tools, which allows users to explore rulings regarding admissibility of expert witnesses’ testimony and reports for product liability and torts. Entering the name of a particular expert witness allows the user to view the admissibility records, including which judges have ruled on the admissibility, who the top parties are that retained the expert and any relevant admissibility order. By searching by judge or court, you can get analytics about the admissibility, top witnesses and top parties who retained an expert witness.
A great attorney is familiar with the law. A great lawyer knows the judge. Lex Machina is a must-know for any modern lawyer.
Frederick’s Crystal Ball
Lex Machina, or its competitor, will continue to develop until AI is able to calculate if a general attorney should pursue trial strategy or settle. The AI will perform a predictive analysis to determine the minimum and maximum settlement ranges.
The AI will consider the following factors when deciding whether trial is the right strategy: the facts, precedents and judge’s past experience in similar cases; which firms performed best; professional witnesses to choose, as well as the demographics of local juries. The AI will provide an analysis on what evidence should be used and not used, preventing litigators from revealing their strategy.
AI can guide or generate everything, from motions and evidence to witnesses to final summations.
Do you mind if I bring a lie detector to the deposition?
Imagine asking a judge to bring in a lie detector during a deposition. You would be laughed out of the office. Now you don’t need to. Depo-IQ takes care of it.
Not exactly. It uses facial recognition software to analyze your video depositions. It can analyze 80 hours of depositions within two hours.
Allen Young is CEO of Depo IQ. “Our AI takes a massive, tens-of-thousands model of human behavior collected from a large population of people in different situations. Health/clinical scenarios are a small minority.” He explained. We build a baseline for the deponent, and compare their responses to every question in a deposition with our model. (Our model is constantly learning and has its own baseline). I believe we use data in order to confirm or deny an attorney’s gut feel about a witness.”
Mr. Young was open about the strengths and weaknesses DepoIQ.
“Our AI is much slower than a lawyer with years of experience who has gained their judgment through court battles.” A machine cannot make a snap decision as fast as an expert for one question. In most fields, experts and laymen have a 50% accuracy rate, while laymen only get it about half the time. That’s a 10% difference between the two. The longer the deposition lasts, the more accurate we become. We can also review hundreds of questions faster than a lawyer expert. We can review a deposition lasting eight hours with hundreds of questions in less than 10 minutes.
While AI cannot replace gut instinct for a single question, the program can review hours of depositions to give attorneys a roadmap for where they may want to look in these hours and use their instinct. I asked him about the reliability.
This technology has been used in hundreds of thousands job interviews, interrogations, and clinical trials. This technology is used by Fortune 500 companies to make hiring decisions that are worth hundreds of thousands dollars every day. This technology is being applied to depositions and we will continue to improve in terms of accuracy. Currently, we are 100% accurate when detecting deviations from baseline behaviors. We are behind in understanding the consequences of someone’s deviation from baseline behavior. “We will figure out quickly how to make the legal industry super reliable.”
I asked if Depo IQ understands questions or just looks at the faces of people to see if they are lying. I was surprised by his answer.
The AI model uses several NLP models to understand the meaning of words. We can surface non-obvious and interesting insights if the lead litigator asks the same questions to each deponent.
He went on to say that Depo-IQ time stamps each place where deception is occurring.
We score every answer using an ‘Alignment Scoring’. The PDF generated shows the scores based on the transcripts and the corresponding deposition videos.
Fredericks Crystal Ball
Depo IQ becomes exponentially more accurate over time. It will ask targeted questions to determine if multiple deponents are lying on the same subject. In court, the attorney could ask questions about that topic to every witness by saying:
Perjury is punishable by jail time. Do you know of any explicit pictures …”
As a chess-player, I’d have my witnesses enter depositions by twitching their faces, blinking rapidly or averting their eyes, etc. So as to send the opposition in the opposite direction.
It will be fascinating to see how AI counters that.
I will be interviewing attorneys and in-house lawyers about AI for my next article. We will discuss AI, including its successes and failures. We’ll also explore why some lawyers are adamantly opposed to it.
The post Tidal Waves of AI Coming to the Legal Profession Part 2 first appeared on Attorney at Law Magazine.