By Carole J. Buckner

Carole J. Buckner is of counsel to Wilson, Elser, Moskowitz, Edelman & Dicker LLP in their San Diego office and a member of LACBA’s Professional Responsibility and Ethics Committee. She represents lawyers and law firms and serves as an expert witness. She may be reached at carole.buckner@wilsonelser.com. Ethics articles are provided regularly by LACBA’s longstanding Professional Responsibility and Ethics Committee.

Lawyers selecting a jury must do so in an ethical, non-discriminatory manner. Historically, this has been highly problematic and it will face greater complications with recent technical developments of   artificial intelligence (AI) software for use in the jury selection process. Given these new technological advancements, recent ABA Formal Opinion 517[1] reinforces the lawyer’s ethical duty to avoid discrimination in jury selection and provides basic guidance on the use of artificial intelligence software in the jury selection process. The use of artificial intelligence (AI) in the practice of law promises many benefits, including a range of AI applications supporting lawyers in jury selection. However, lawyers must proceed with caution to ensure compliance with their ethical duties.      

Under Batson v. Kentucky, 476 US 79 (1986) and its progeny, trial lawyers are prohibited from using peremptory challenges based on race or gender, because doing so violates potential jurors’ constitutional rights to equal protection, and can also violate state law.  Rule 8.4 of the Model Rules of Professional Conduct restricts lawyers from conduct that a lawyer “knows or reasonably should know is harassment or discrimination on the basis of race, sex, religion, national origin, ethnicity, disability, age, sexual orientation, gender identity, marital status or socioeconomic status in conduct related to the practice of law.”[2] Similarly, California’s Rule 8.4.1(1) provides that “in representing a client, … a lawyer shall not … unlawfully discriminate against persons on the basis of any protected characteristic.”[3] This rule regulates conduct of California lawyers in proceedings before judicial officers, but a finding that peremptory challenges were exercised in a discriminatory manner does not alone establish a disciplinary violation.[4] The rule also provides that lawyer may not engage in such conduct through the “acts of another.”[5]

California lawyers must also comply with applicable statutes governing the use of peremptory challenges enacted to address the ineffectiveness of Batson in preventing discriminatory peremptory strikes, including implicit bias.[6]  California lawyers may exercise peremptory challenges without stating a reason, but may not exercise peremptory challenges to remove a prospective juror on the basis of the prospective juror’s race, ethnicity, gender, gender identity, sexual orientation, national origin, or religious affiliation, or the perceived membership of the prospective juror in any of those groups.[7] The opposing party may object to the use of a peremptory challenge, in which case the party exercising the peremptory challenge must state the reasons for the peremptory challenge, which the court must then evaluate to determine whether there is a substantial likelihood that an objectively reasonable person would view a protected classification as a factor in the use of the peremptory challenge.[8] For purposes of judicial evaluation, an objectively reasonable person is considered to be aware that unconscious bias (defined as implicit and institutional bias), as well as purposeful discrimination, can result in the unfair exclusion of jurors.[9]

The court may consider a wide variety of circumstances in its evaluation of how an objectively reasonable person would view the challenge, including many nuances of the voir dire questioning.[10]  Peremptory challenges for a host of specified reasons, such as the juror’s neighborhood, lack of employment, dress, attire or personal appearance, or field of employment, are presumed invalid unless the party exercising the peremptory challenge can show by clear and convincing evidence that an objectively reasonable person would view the rationale as unrelated to a cognizable group.[11] Any justification applicable to prospective jurors who are not members of a perceived cognizable group but were not challenged is also presumed invalid unless the required showing is made.[12]

Opinion 517 points out that when lawyers exercise the peremptory challenge themselves, they “know” the real reasons for those challenges. But lawyers may rely on others, including clients or jury consultants, in deciding how to exercise peremptory challenges. When a lawyer defers to others, whether the lawyer is violating the ethical prohibition against discrimination may be a more difficult question.[13]  For example, when a lawyer is working on jury selection with a client or a jury consultant who is indicating that the reason for the peremptory is race or gender based, a lawyer accepting such an instruction or suggestion would be “knowingly” discriminating, in violation of the rule.[14] However, the client or jury consultant may provide sincere, non-discriminatory reasons for the exercise of peremptory challenges, which a lawyer may advance only if a competent, reasonably prudent lawyer would not find them pretextual, in which case the lawyer must refrain from advancing them.[15] Opinion 517 also points out the lawyer’s duty to consult with the client, including explaining relevant constraints pertaining to jury selection.[16]

Opinion 517 explains that analogous issues can arise when a lawyer relies on software to assist the lawyer with jury selection.[17] A variety of artificial intelligence assisted software services are available to assist with various aspects of jury selection.[18] Some software services facilitate data gathering and organization of information regarding the prospective jurors, while other technology uses a scoring algorithm to generate rankings of potential jurors.[19] AI systems can research juror information online, analyze data from jury questionnaires, gather social media, then leverage machine learning to analyze the data, including “highlighting potential biases.”[20] Other AI jury selection software reportedly evaluates, scores, and ranks prospective jurors.[21]

Opinion 517 posits that an artificial intelligence assisted program could rank prospective jurors in a manner that would constitute unlawful discrimination while providing seemingly neutral reasons for the rankings.[22] In such a situation the Opinion concludes that a lawyer using the program would engage in purposeful discrimination in violation of Batson, even if the lawyer had no intention of doing so. The question is whether the lawyer “reasonably should know” that the peremptory challenges were impermissibly discriminatory.[23]  The Opinion indicates that this will depend on the circumstances.

Generative artificial intelligence is in general fraught with bias.[24] California State Bar guidelines warn that some generative artificial intelligence is trained on biased information.[25] Compliance with rule 8.4.1 requires that lawyers be aware of possible biases, and engage in continuous learning about AI biases, and their implications in legal practice.[26]

Opinion 517 provides that lawyers have a “duty of inquiry” to engage in due diligence to acquire an understanding of the methodology used by jury selection software, and to understand the benefits and risks, capabilities and limitations.[27] Lawyers may do this through self-study, by associating another competent lawyer, or consulting with another person with expertise in the field.[28] To provide competent representation, a lawyer must engage in an appropriate amount of independent verification and review of the output to avoid misleading representations to the court.[29]

Lawyers can read about AI tools, and attend continuing legal education.[30] In utilizing AI jury selection programs that evaluate and rank prospective jurors, lawyers must inquire regarding whether jury selection software bases its algorithm on race, sex, religion, national origin, ethnicity, disability, age, sexual orientation, gender identity, marital status, or social economic status, in order to avoid using artificial intelligence algorithms that analyze protected characteristics to predict bias. For California lawyers, a more nuanced inquiry compliant with detailed statutory requirements will also be relevant. In addition, although independent verification in this context may be difficult, lawyers using jury selection ranking programs must examine the output for inappropriate bias.

The duty to understand the technology a lawyer is using is nothing new, however, gaining an understanding of artificial intelligence poses some distinct challenges.  AI systems can autonomously reconfigure their controls, take actions contrary to their training, and misrepresent that they have done so.[31] AI systems change over time, generating varying output.[32] Transparency with AI systems is challenging given that they can be a “black box” even to their developers.[33]

Opinion 517 imposes an ethical duty to “conduct sufficient due diligence to acquire a general understanding of the methodology employed by the juror selection program.”[34] This itself may pose a problem if the methodology is a proprietary algorithm such that transparency is lacking. The ability of AI to conceal its own bias presents further issues, as does the evolving nature of AI technology, raising questions about what lawyers can truly know about the methodology used. We know from numerous cases that AI can be totally wrong about the law.[35]

At the moment, whether the use of AI tools in jury selection provides a strategic advantage is open to debate.[36] Yet, lawyers have a duty to keep abreast of the changes in the law and its practice, including the benefits and risks associated with relevant technology.[37] Lawyers have a duty to understand and properly utilize technology in their practice, such that, in the near future, the practice of using AI to help with jury selection may be viewed as another aspect of competent lawyering.[38] Opinion 517 points out that lawyers using software employing AI in jury selection must have a reasonable understanding of how it functions in order to assure that they meet their ethical duties.