This pilot study evaluates three machine learning models—Claude, ChatGPT, and Gemini—in parole decision-making by comparing their recommendations against human judicial decisions. Analyzing 150 cases across sexual, drug-related, and violent offenses, the study demonstrates alignment rates of 56%–68% between artificial intelligence (AI)-generated and human decisions. Findings reveal a consistent rehabilitative bias in AI recommendations, with statistical analyses indicating robust patterns across offense categories. This research establishes an empirical foundation for understanding AI’s potential as a decision-support tool in parole processes while preserving essential human judgment. While focusing on Israel, the findings offer a preliminary basis for considering similar AI integrations across different jurisdictions. These results underscore the importance of developing clear criminal justice policies and ethical frameworks to guide the responsible use of AI in parole decisions, ensuring that technological tools support rather than replace human judgment and enhance public trust in the system.
ArowosegbeJ. O. (2023). Data bias, intelligent systems, and criminal justice outcomes. International Journal of Law and Information Technology, 31(1), 22–45.
6.
BarocasS.SelbstA. D. (2016). Big data’s disparate impact. California Law Review, 104(3), 671–732. https://doi.org/10.15779/Z38BG31
7.
BerkR.HeidariH.JabbariS.KearnsM.RothA. (2021). Fairness in criminal justice risk assessments: The state of the art. Sociological Methods & Research, 50(1), 3–44. https://doi.org/10.1177/0049124118782533
8.
BinnsR. (2018). Fairness in machine learning: Lessons from political philosophy. Proceedings of Machine Learning Research, 81, 149–159.
9.
DaganN. (2023). Parole as resentencing: Exploring the punitive accounts of parole decision-making through the comparative case study of Israel. European Journal of Criminology, 20(4), 1231–1250. https://doi.org/10.1177/14773708211039635
10.
DaganN.TimorA.RonelN. (2013). Parole release objectives as viewed by parole board heads and their discretion in their application. Tzohar LeBeit HaSohar, 15, 62–81.
11.
DouL.DouX. (2025). Towards just AI: Challenges and solution framework for algorithmic discrimination in judicial systems. International Journal of Digital Law and Governance. Advance online publication. https://doi.org/10.1515/ijdlg-2024-0020
12.
EloS.KyngäsH. (2008). The qualitative content analysis process. Journal of Advanced Nursing, 62(1), 107–115.
13.
FarayolaM. M.TalI.BendechacheM.SaberT.ConnollyR. (2023). Fairness of AI in predicting the risk of recidivism: Review and phase mapping of AI fairness techniques. In Proceedings of the 18th International Conference on Availability, Reliability and Security (ARES 2023) (pp. 1–10). ACM. https://doi.org/10.1145/3600160.3605033
14.
FarberS. (2024). AI in terrorism sentencing: Evaluating predictive accuracy and ethical implications. Criminal Justice Studies, 37(4), 352–370. https://doi.org/10.1080/1478601X.2024.2415558
15.
FarberS. (2025). AI as a decision support tool in forensic image analysis: A pilot study on integrating large language models into crime scene investigation workflows. Journal of Forensic Sciences. Advance online publication. https://doi.org/10.1111/1556-4029.70035
16.
GanguliD.HernandezD.LovittL.AskellA.BaiY.ChenA.ConerlyT.DassarmaN.DrainD.ElhageN.El ShowkS.FortS.Hatfield-DoddsZ.HenighanT.JohnstonS.JonesA.JosephN.KernianJ.KravecS.. . .ClarkJ. (2022, June). Predictability and surprise in large generative models. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency (pp. 1747–1764). ACM.
JavedK.LiJ. (2025). Bias in adjudication: Investigating the impact of artificial intelligence, media, financial, and legal institutions in pursuit of social justice. PLOS ONE, 20(1), Article e0315270.
23.
JiangC.YangX. (2024). Agents on the Bench: Large language model-based multi-agent framework for trustworthy digital justice. arXiv. https://arxiv.org/abs/2412.18697
24.
KatzD. M.BommaritoM. J.BlackmanJ. (2017). A general approach for predicting the behavior of the United States Supreme Court. PLOS ONE, 12(4), Article e0174698.
25.
KleinbergJ.LakkarajuH.LeskovecJ.LudwigJ.MullainathanS. (2018). Human decisions and machine predictions. The Quarterly Journal of Economics, 133(1), 237–293. https://doi.org/10.1093/qje/qjx032
26.
KrippendorffK. (2018). Content analysis: An introduction to its methodology (4th ed.). Sage.
27.
LandauS.DempseyJ. X.KamarE.BellovinS. M.PoolR. (2024). Challenging the machine: Contestability in government AI systems. arXiv preprint arXiv:2406.10430
28.
LaqueurH. S.CopusR. W. (2024). An algorithmic assessment of parole decisions. Journal of Quantitative Criminology, 40(1), 151–188.
29.
LincolnY. S.GubaE. G. (1985). Naturalistic inquiry. Sage.
30.
MarchJ. G.SimonH. A. (1993). Organizations. Wiley-Blackwell.
31.
MehrabiN.MorstatterF.SaxenaN.LermanK.GalstyanA. (2021). A survey on bias and fairness in machine learning. ACM Computing Surveys, 54(6), Article 115. https://doi.org/10.1145/3457607
32.
O’NeilC. (2016). Weapons of math destruction: How big data increases inequality and threatens democracy. Crown. https://doi.org/10.5860/crl.78.3.403
RybergJ. (2024). Criminal justice and artificial intelligence: How should we assess the performance of sentencing algorithms?Philosophy & Technology, 37(1), 9.
36.
RybergJ. (2025). Artificial intelligence and criminal justice: How to use algorithmic sentencing support in real-life (and ethically non-ideal) penal systems?AI and Ethics. Advance online publication. https://doi.org/10.1007/s43681-024-00655-8
37.
StarrS. B. (2014). Evidence-based sentencing and the scientific rationalization of discrimination. Stanford Law Review, 66, 803–872.
TaylorI. (2023). Justice by algorithm: The limits of AI in criminal sentencing. Criminal Justice Ethics, 42(3), 193–213.
40.
U.S. Department of Justice, Office of Legal Policy. (2024). Artificial intelligence and criminal justice: Final report. U.S. Government Publishing Office.
