Asian Journal of Criminal Justice and Forensic Studies

  • Home
  • Vol. 1, No. 1. 2025

  • The implementation of artificial intelligence in Singapore’s legal practice: Potential and risks for criminal investigations

  • Received 23.06.2025,
  • Revised 10.11.2025,
  • Accepted 16.12.2025
Download article Download article
Vol. 1, No. 1. 2025
  • reliability; algorithm; reproducibility; data; verification; triage
  • Pages 62-74

The aim of this study was to assess the conditions for applying algorithmic systems in the activities of law enforcement and judicial oversight bodies in Singapore, in order to determine the boundaries of their admissibility with regard to the guarantees of due process. The methodology was based on normative-analytical, risk-oriented, structural-logical modelling, analytical synthesis, and case-study methods. It has been established that Artificial Intelligence automates cognitive tasks (classification, prediction); however, the results of machine learning are probabilistic and require regular quality testing and monitoring. In practice, AI analytics primarily generate “candidate” signals (anomalies, risk rankings) that require independent confirmation. The evidentiary status of digital traces arises after the provenance is recorded, and the integrity and verifiability of the materials are ensured. In financial crimes, technologies scale up both investigations and prevention, but automation creates a “risk paradox”: errors in data or settings scale as rapidly as the positive effects of the system. It was found that Singapore’s courts permit the use of Generative Artificial Intelligence only as an ancillary tool, placing full responsibility on the user for the accuracy and correctness of the submitted materials. The use of the Scam Analysis and Tactical Intervention System Plus enabled the triage of 7,200 monikers and the issuance of 3,700 directions under the Online Criminal Harms Act. Concurrently, the implementation of the Automation of Scam-fighting Tactics & Reaching Out ensured the automation of SMS alert dissemination, which helped prevent losses amounting to $420.41 million. At the same time, judicial practice indicates that operational effectiveness is not synonymous with evidentiary admissibility, as the authenticity of data requires a separate justification of the algorithmic output’s reliability. To minimise procedural risks, it is advisable to apply standardised mechanisms for logging, model version control, error documentation, and independent verification. The practical significance lies in the application of the results by law enforcement agencies, prosecutors, courts, and the defence in Singapore when evaluating and using algorithmic results in criminal proceedings

References

  1. AI Verify Foundation. (2024). Proposed Model AI Governance Framework for Generative AI: Fostering a trusted ecosystem. Retrieved from https://aiverifyfoundation.sg/downloads/Proposed_MGF_Gen_AI_2024.pdf.
  2. Alibašić, H. (2025). Harmonizing artificial intelligence (AI) governance: A comparative analysis of Singapore and France’s AI policies and the influence of international organizations. Global Public Policy and Governance, 5, 93-113. doi: 10.1007/s43508-025-00116-w.
  3. Allen, J.G., Loo, J., & Luna, J.L. (2025). Governing intelligence: Singapore’s evolving AI governance framework. Cambridge Forum on AI: Law and Governance, 1, article number e12. doi: 10.1017/cfl.2024.12.
  4. Almasoud, A.S., & Idowu, J.A. (2024). Algorithmic fairness in predictive policing. AI and Ethics, 5, 2323-2337. doi: 10.1007/s43681-024-00541-3.
  5. Beltrão, G., Goh, S.T., Sousa, S., & Lamas, D. (2025). Community, identity & stability? Building trust in facial recognition systems for mass surveillance. Journal of Responsible Technology, 24, article number 100139. doi: 10.1016/j.jrt.2025.100139.
  6. Berk, R.A., Kuchibhotla, A.K., & Tchetgen Tchetgen, E. (2021). Improving fairness in criminal justice algorithmic risk assessments using optimal transport and conformal prediction sets. ArXiv. doi: 10.48550/arXiv.2111.09211.
  7. Chase, J., Phong, T., Long, K., Le, T., & Lau, H.C. (2021). Grand-vision: An intelligent system for optimized deployment scheduling of law enforcement agents. Proceedings of the International Conference on Automated Planning and Scheduling, 31(1), 459-467. doi: 10.1609/icaps.v31i1.15992.
  8. Court of Appeal No. 42 “GIL v Public Prosecutor”. (2025, May). Retrieved from https://www.elitigation.sg/gdviewer/s/2025_SGCA_21.
  9. Douglas, T., Davies, B., Pugh, J., Brown, R., Hass, B., Forsberg, L., Mishra, A., Singh, I., Savulescu, J., & Fazel, S. (2021). Algorithmic risk assessment tools in criminal justice: The need for better data. Oxford: University of Oxford.
  10. Dunsin, D., Ghanem, M.C., Ouazzane, K., & Vassilev, V. (2024). A comprehensive analysis of the role of artificial intelligence and machine learning in modern digital forensics and incident response. Forensic Science International: Digital Investigation, 48, article number 301675. doi: 10.1016/j.fsidi.2023.301675.
  11. Evidence Act of the Republic of Singapore. (1997, December). Retrieved from https://sso.agc.gov.sg/Act-Rev/97/Published?DocDate=19971220&ProvIds=pr76-.
  12. Greene, T., Shmueli, G., Fell, J., Lin, C.-F., & Liu, H.-W. (2022). Forks over knives: Predictive inconsistency in criminal justice algorithmic risk assessment tools. Journal of the Royal Statistical Society: Series A (Statistics in Society), 185(2), 692-723. doi: 10.1111/rssa.12966.
  13. Guide on the Use of Generative Artificial Intelligence Tools by Court Users. (2024). Retrieved from https://www.judiciary.gov.sg/docs/default-source/news-and-resources-docs/guide-on-the-use-of-generative-ai-tools-by-court-users.pdf?sfvrsn=3900c814_1&.
  14. Home Team Science and Technology Agency. (2025). Digital & information forensics. Retrieved from https://www.htx.gov.sg/who-we-are/what-we-do/our-expertise/digital-information-forensics.
  15. Infocomm Media Development Authority. (2024). Singapore proposes framework to foster trusted Generative AI development. Retrieved from https://www.imda.gov.sg/resources/press-releases-factsheets-and-speeches/press-releases/2024/public-consult-model-ai-governance-framework-genai.
  16. Khan, A.A. (2024). Reconceptualizing policing for cybercrime: Perspectives from Singapore. Laws, 13(4), article number 44. doi: 10.3390/laws13040044.
  17. Lee, H., & Lee, T. (2022). The tracetogether matrix has you – surveillance, rationalisation and tactics of governance in Singapore’s COVID-19 app. Platform: Journal of Media and Communication, 9(2), 77-91.
  18. Lee, T., & Lee, H. (2020). Tracing surveillance and auto-regulation in Singapore: ‘Smart’ responses to COVID-19. Media International Australia, 177(1), 47-60. doi: 10.1177/1329878X20949545.
  19. Lettieri, N., Guarino, A., Zaccagnino, R., & Malandrino, D. (2023). Keeping judges in the loop: A human-machine collaboration strategy against the blind spots of AI in criminal justice. Soft Computing, 27, 11275-11293. doi: 10.1007/s00500-023-08604-z.
  20. Lim, N.F.-Z., & Tan, K.S. (2025). The new frontier: Regulating artificial intelligence in Singapore. Singapore Academy of Law Journal, 37, 436-463.
  21. Lim, S.S., & Chng, G. (2024). Verifying AI: Will Singapore’s experiment with AI governance set the benchmark? Communication Research and Practice, 10(3), 297-306. doi: 10.1080/22041451.2024.2346416.
  22. Loovens, J., & Tinmaz, H. (2025). A systematic literature review of deepfakes in forensic science. Forensic Imaging, 43, article number 200647. doi: 10.1016/j.fri.2025.200647.
  23. Magistrate’s Appeal No. 9043 “GIL v Public Prosecutor”. (2024, November). Retrieved from https://www.elitigation.sg/gd/s/2024_SGHC_287.
  24. Maoro, F., & Geierhos, M. (2025). Contestable AI for criminal intelligence analysis: Improving decision-making through semantic modeling and human oversight. Frontiers in Artificial Intelligence, 8, article number 1602998.  doi: 10.3389/frai.2025.1602998.
  25. Originating Claim No. 125 (Summons No. 1240 of 2025) “Tajudin bin Gulam Rasul and Mohamed Ghouse v. Suriaya bte Haja Mohideen”. (2025, September). Retrieved from https://www.elitigation.sg/gd/s/2025_SGHCR_33.
  26. Personal Data Protection Commission. (2020). Model AI governance framework. Retrieved from https://www.pdpc.gov.sg/-/media/files/pdpc/pdf-files/resource-for-organisation/ai/sgmodelaigovframework2.pdf.
  27. Purves, D. (2022). Fairness in algorithmic policing. Journal of the American Philosophical Association, 8(4), 741-761. doi: 10.1017/apa.2021.39.
  28. Qureshi, S.M., Saeed, A., Almotiri, S.H., Ahmad, F., & Al Ghamdi, M.A. (2024). Deepfake forensics: A survey of digital forensic methods for multimodal deepfake identification on social media. PeerJ Computer Science, 10, article number e2037. doi: 10.7717/peerj-cs.2037.
  29. Sandoval, M.-P., Vau, M. de A., Solaas, J., & Rodrigues, L. (2024). Threat of deepfakes to the criminal justice system: A systematic review. Crime Science, 13, article number 41. doi: 10.1186/s40163-024-00239-1.
  30. Seng, D., & Mason, S. (2021). Artificial intelligence and evidence. Singapore Academy of Law Journal, 33, 241-279.
  31. Singapore Courts. (2024). Strengthening justice, safeguarding society. In Singapore courts annual report 2023. Singapore: The Judiciary.
  32. Singapore Police Force. (2024a). CAD report 2024. Retrieved from https://www.police.gov.sg/-/media/SPF/Media-Room/Publications/CAD-Report-2024/CAD-Report-2024.pdf.
  33. Singapore Police Force. (2024b). A future-ready Singapore Police Force: Cyber and beyond. Retrieved from https://www.police.gov.sg/-/media/SPF/Files/Publications/PDF/SPF-Annual-Report-2024.pdf.

Submission