Exploring the use of psychographic profiling and neurolinguistic AI tools in the legal system

Share on facebook
Share on twitter
Share on linkedin
Share on reddit
Share on telegram
Table of Contents

In the Sixth Amendment in the United States’ Bill of Rights, criminal defendants are guaranteed certain rights, including the right to an impartial jury. However, in today’s age of info-tainment, the never-ending news cycles bombard potential jurors, judges, arbitrators, and other judicial actors with information—or misinformation—that can bias their decision-making. Consider the Johnny Depp defamation suit against his ex-wife, Amber Heard—two famous celebrities with their own legion of adoring fans locked into a very public trial that has a lot of media attention. How does a juror maintain impartiality when they are bombarded with media opinions and stories like Amber Heard’s IMDB page being changed to “Amber Turd”? This begs the question further, can we even hope to find an impartial juror in the age of social media?

It’s a stunning question that most people would shy away from considering. Most courtroom cases do not involve celebrities or other famous people, but consider how “Google-able” each one of us is to a potential juror. Some argue that jury questionnaires are specifically designed to weed out biased juror candidates. For instance, the OJ Simpson trial had a seventy-five page questionnaire because of the famous nature of the defendant. If you read through it, it does ask questions like have you ever seen OJ play football? Or as an actor? Spokesperson? How about the question: Do you follow football? Do you like or hate the University of Southern California? The likelihood is that we’ll find 12 to 15 people who have not been influenced in some capacity seems unlikely.

So, what can we do about the rising challenges bias presents in the face of the fair decisions judicial proceedings are entrusted to bring? A few futurists have floated the idea of artificial intelligence (AI) jurors or even AI robot judges to ensure impartiality. While these AI bots could be immune to the influence of media stories, there are implicit bias considerations in how they were trained that would need to be addressed. (Yes, even our training data can be biased so machines are not a magic bullet solution). There is hope, however, with some AI tools for our legal and judicial teams. Tapping into the hard sciences of psychographics and neurolinguistics, we can better identify biases in potential jurors and monitor them throughout the trial to see if non-courtroom information may be unduly swaying their opinion.

Psychographics is the science of understanding people according to their hobbies, interests, and attitudes. In the world of AI, many tools already exist to generate psychographic profiles of an individual. Using just public information like social media posts, LinkedIn biographies, and so forth, the AI has enough data to generate personality trait scores—across all 56 traits—and decipher the opinions, attitudes, hobbies, and even political party affiliation of a person. Thus, at an individualized level, AI can decode a person.

Psychographics is then complemented by neurolinguistics, which is the science of how language is represented within the brain. In essence, language is like a fingerprint and reveals clues to our values, commitment level, and beliefs. Moreover, it is very difficult to disguise. With enough verbal or written data, the “true person” is always revealed. There are already a few companies that have developed AI tools using neurolinguistics to analyze legal depositions, help depressed teenagers, and augment individualized marketing messages.

Let’s revisit the Johnny Depp defamation trial assuming we had these psychographic profiling and neurolinguistic AI tools. From the questionnaire data alone, the AI would have enough data to assess each potential juror. Now, imagine knowing—in advance—how each person truly feels about Johnny Depp and Amber Heard, defamation cases, and other potential biases. In addition, we would know which media channels each prospective juror is likely to engage with, which would help us assess the probability of media persuasion shaping this person’s opinions. Furthermore, for those selected as jurors, we have a baseline on their impartiality. As the case progresses, we can reassess their psychographics and neurolinguistics to see if outside forces are impacting the juror’s perception of the case. Thus, even with intense media coverage—and both legal teams trying to shape that coverage for their client’s benefit—these AI tools provide the ability to better sustain an impartial jury.

Some people will argue about data security and privacy rights about using this information from a potential juror. Given the level of personal intimacy required from these questionnaires, though, this argument does not seem to hold valid, especially when people are often expected to answer some of these questions publicly during jury selection. There is another drawback here that does require serious consideration. If these AI tools are effective in screening out biased juror candidates, would we be left with a large enough pool of people to select from? In most non-famous cases, the answer is yes. In celebrity cases like Johnny Depp and Amber Heard, however, that’s a tricky one. It’s difficult to expect that there’s a large enough pool of people out there who have not heard from—especially with the media coverage going on. Would this mean that we would have to stitch a pool of candidates together from across different counties or even states to find enough impartial people? It is possible. Alternatively, the legal system may just have to accept some level of bias to even find enough jurors within the venue.

Ultimately, there is no perfect solution, but that is not the goal. The main focus is to improve our ability to provide an impartial jury, and AI could be a solid helping hand to accomplish this. That is why it is worthwhile to explore the use of psychographic profiling and neurolinguistic AI tools for our legal system.

Share on facebook
Share on twitter
Share on linkedin
Share on reddit
Share on telegram
Neil Sahota
Neil Sahota (萨冠军) is an IBM Master Inventor, United Nations (UN) Artificial Intelligence (AI) Advisor, author of the best-seller Own the AI Revolution and sought-after speaker. With 20+ years of business experience, Neil works to inspire clients and business partners to foster innovation and develop next generation products/solutions powered by AI.