FPSY3900: The Future of Jury Decision-Making

Lecture 10 (Final Testable Lecture)

Culture Differences in Jury Deliberations & Jury Decision Making between South Korea and the United States (Alcohol)

• identify differences between collectivist and individualistic jurors

• communication styles

• how groups’ deliberations differed

• attribution styles; relates to the defendant’s behaviour

  • blame → is attributed to something internal or external

Lee et al. (2025)

Method

• 28 mock jury deliberations (14 Korea/14 US)

• Analysis of transcripts

Findings (collectivist)

• Less conformity (less likely to conform to the group)

• Incorporation of multiple arguments

  • tend to look at more arguments/components to find their final verdict

• Acknowledgment of mitigating factors (external attribution)

  • alcohol abuse

  • more likely to attribute blame to external forces

• No differences in final sentencing 

House of Worship Mass Shooting: The Influence of Defendant Age, Religion and Victim Religion on Mock-Juror Decision-Making (Article)

looked at Christian vs. Muslim defendants, victims

Vettese et al. (2024)

Method

• 321 participants

  • all-white Canadians

• Mock trial transcripts regarding mass shooting at a place of worship

• asked to provided final verdicts

Findings

• Christian defendants = more guilty than Muslim defendants

  • seen less favourably

  • for christians they are seen as more guilty

• Muslim victim = higher guilt ratings

  • more favourably when the defendant is christian

• mock jurors assigned higher guilt in victims

Delayed Reporting & Jury Decision-Making

• don’t know how to report

• may not have an accurate depiction of the events

• feel embarrassed

How Length of & Reason for Delayed Reporting Influence Mock-Jurors’ in a Sexual Assault Trial (Article)

• delaying cases

• fear of retaliation

Thompson & Pozzulo (2024)

Method

• 709 Mock jurors

  • how truthful they thought the testimonies were from both sides

• Delayed reporting and reason for delay

Findings

• Less delay = increase guilty verdict

• Delay reporting due to family concerns = increase guilty verdict

• dichotomos verdict

• men SA/rape the women

• shorter the delay it looked better for the victim

• mock jurors were more likely to render a guilty verdict

• guilty verdict if the victim delayed due to scared for family finding out

Martin & Monds (2023)

Method

• 228 mock jurors

• Rape/robbery transcript

  • committed against a women when they were under the influence

Findings

• Alcohol seen as detrimental to victim credibility

  • do you actually recall the events that occurred

  • intoxication status failed to influence verdict decisions

• No difference in verdict decisions

Threat of pre-trial publicity can threaten blank responses

open-minded

Interventions for the threat of pre-trial publicity

• less biased reporting

• expert testimony

  • to understand the impact of technology

The temptation to seek information

• motivation to be accurate

• case-specific searches

Intervention for the temptation to seek information

restricting access

• when told not to look at the media but they still continue to do so

Asian countries have a more blank notion

interdepent

AI use in court

• Relatively new idea

• Being adopted by some judges

Guidelines for AI use in court

• Canadian Judicial Council

• 7 guidelines for AI use in court

Guideline 1 – Protect judicial independence

• AI use should not be governed by a state agency

• Constitutionally independent courts retain judicial independence

  • government should not be intervening in the ai court systems

  • could play a big part in the courts an d must firmly uhold the idea of independence

  • should not ben e governend by AI structure

• AI applications harbour the potential to erode judicial agency and independence.

Guideline 2 – Use AI consistently with core values and ethical rules

• AI use should align with core values and judicial ethics

• These include independence, integrity and respect; diligence and competence; equality and impartiality, fairness, transparency, accessibility, timeliness and certainty

• AI can remove core ethical values, so we should not let it occur

Guideline 3 – Have regard to the legal aspects of AI use

• Courts should be aware of the source material used to train AI systems

• Consideration for the legality of generative AI practices

• sensitive data are being sent to the courts; giving AI this information

Guideline 4 – AI tools must be subject to security standards 

• AI tools bring unique security challenges

• Robust information and cybersecurity program

• Target security of information and tampering

• unauthorized data

• data is vulanerbale, we need to protect AI components from tamepring

• ensuring data protection

Guideline 5 – AI tools must provide understandable explanations for decision-making 

• AI can be used to improve court efficiency

• Must ensure the explainability of AI systems

• provide lcear outputs

• should be easy for us to interpret to test outputs in courts

Guideline 6 – Courts must track AI impact 

• Administrators must perform a comprehensive, formal, and impartial assessment of its impact

• This relates to judicial independence, workload, backlog reduction, privacy, security, access to justice, and the court’s reputation

Guideline 7 – Program of education and user support  

• Judiciary must have appropriate knowledge of AI systems

• AI should not be employed without users undergoing a comprehensive educational process

• should not be tampered with

• judges/court personnel must be trained to work alongside AI

• to help make sure the courts are running smoothly

AI in the Courtroom: In class video

robot is acting as a judge

AI tools in court

Significant development in recent years

What was the traditional jury shortfalls?

• backlog of cases

  • takes way too long; years

• consistency of verdicts and hung juries

  • juries can be divided in verdicts

• Unreasoned verdicts and witness evaluation 

How would AI help the backlog of cases?

• Decisions 7x faster

• 24h juries

How would AI help the consistency of verdicts and hung juries?

• Reliance on code and algorithm (not values and beliefs)

• Objective and consistent

  • no personal biases

  • remove voire dire process

How would AI help the unreasoned verdicts and witness evaluation?

• Detection of behaviour/reliability 

• Consistent observation of ‘live evidence’

• program AI to do reasonings

• no issues in interpreting legal proceedings

Potential limitations/risks of AI

• Jury tampering

• Bias

• Lack of conscience