Authored By: Maria Shanga Masseka
Maria Shanga Masseka
1 INTRODUCTION
Artificial intelligence (AI) refers to the ability of a computer-controlled robot to perform tasks that are typically associated with the intellectual processes characteristic of humans, such as reasoning. Artificial intelligence affects on the legal system, begun by reshaping how evidence is collected, analyzed, and presented in court.
Courts normally have to make a finding concerning the existence or non-existence of certain facts before pronouncing on the right duties and liabilities of parties engaged in a dispute. In the process of litigation and adjudication, the proof of facts is regulated by the law of evidence, which is a branch of the law of procedure that ensures justice and a fair trial by providing general rules and regulations regarding the admissibility of different types of evidence.
As Artificial evidence has become more common in the 21st century, traditional evidentiary principles such as reliability are being tested. Therefore, this article examines the growing impact of Artificial Intelligence on the law of evidence by exploring both the benefits and challenges it introduces. Although Artificial intelligence does enhance efficiency and accuracy in handling evidence, it does raise complex issues regarding the reliability, transparency, and accountability of traditional evidentiary rules.
2 OVERVIEW OF LAW OF EVIDENCE
The main function of the law of evidence is to determine what facts are legally admissible to prove the facts in issue. Law of evidence however also determines in what manner evidence should or maybe adduced, what evidence may lawfully be withheld from court of law what rules should be taken into account in assessing the weight or cogency of evidence and further, what standards of proof should, in a given situation, be satisfied before a party bearing the burden of proof can be successful.1 There is no degree of admissibility of evidence is either admissible or inadmissible.2 Evidence cannot be more admissible or less admissible, however, once it is admissible it may carry more or less weight according to the particular circumstance of the case.3 The courts therefore evaluates or weigh evidence to determine whether the required standard of proof has been attained, which for criminal law procedure is beyond a reasonable doubt and for civil law procedure is on balance of probability in terms of the common law principles.
2 1 ADMISSIBILTY OF EVIDENCE
Admissibility of evidence is a principle determined with reference to its relevance, which is done so with reference of the necessity also made to the potential weight of the evidence.4 Section 210 of the criminal Law Procedure Act provides that no evidence as to any fact, matter or anything shall be admissible if irrelevant or immaterial and if it cannot conduce to prove or disprove any point or provision.5 These section serves as a statutory confirmation of the South African common law and state the rule in its negative form irrelevant evidence is inadmissible, however courts are inclined to states the rule in its positive form all facts relevant to the issue in legal proceeding maybe proved.6In the case of S v Zuma, Van der Merwe J said “The question of relevancy can never be divorced from the facts of a particular case court.”7 Evidence is divided into categories such as oral evidence (witness testimony), documentary evidence, and real evidence (physical objects), with the digital revolution, electronics, and digital evidence such as emails, digital images, and electronic transactions.
UNDERSTANDING ARTIFICIAL INTELLIGENCE IN THE LEGAL CONTEXT
Artificial Intelligence refers to computer systems capable of performing tasks that typically require human intelligence, such as reasoning, learning, and decision-making. In legal practice, AI is increasingly used in document review, e-discovery, legal research, predictive analysis, and digital forensics.8(Ashley,2017). Furthermore, Artificial intelligence has enormous potential in legal field, promising to transform how legal practitioners function and individuals access the justice system. The function of AI in the legal domain is diverse AI-driven research tools can rapidly analyze extensive legal databases, furnishing attorneys with insights in a previously unattainable manner.9 AI in law of evidence it entails the utilization of chatbots and virtual assistance to provide individuals with legal guidance, thereby enhancing the accessibility of legal information. 10 While these technologies improve efficiency, they raise questions about transparency, bias, and accountability, particularly when AI findings influence judicial decisions.
3 THE IMPACT OF ARTIFICIAL INTELLIGENCE ON EVIDENCE LAW 3 1 Collection of Evidence
Artificial Intelligence facilitates the collection of numerous data from multiple sources such as social media, surveillance cameras, and digital devices, which falls under documentary evidence. Tools like predictive policing algorithms assist law enforcement in identifying potential suspects or locations of crime.11 (Ferguson, 2017). However, concerns arise regarding the authenticity and chain of custody of AI-collected evidence. If an AI tool autonomously gathers or processes data, questions emerge about whether the evidence was tampered with or whether the tool operated correctly. Moreover, in the article of Don’t rely on Bots magistrate warns it reports that in a cautionary tale, a South African law has been outed for using ChatGPT to research a case law, which threw up cases that did not exist.12 In the case of Mavundla v MEC: Department of Cooperative Government and Traditional Affairs KwaZulu-Natal and Others it was emphasized that courts expect lawyers to bring a legally independent and questioning mind to bear on, especially, novel legal matters.13 Reliance on AI technologies when doing legal research is irresponsible and unprofessional.14 The judgment of the case was referred to the Legal Practice Council.15
3 2 Analysis and Evaluation of Evidence
AI can analyze complex evidence, such as DNA samples, fingerprints, or digital communications, far more quickly and accurately than humans. Forensic AI tools can identify forged documents or detect inconsistencies in witness statements.16 Nonetheless, the real problem is that the inability to explain in more detail how AI reaches its conclusions poses challenges for cross-examination and judicial understanding.
3 1 3 Admissibility and Weight of AI Evidence
Courts are increasingly confronted with questions about whether AI-generated evidence meets the standards of admissibility. Evidence law requires that the source and process of evidence be reliable, transparent, and authentic. If the algorithm used to produce the evidence is proprietary or secret, it becomes difficult for courts to assess reliability. Expert testimony is often required to explain how AI systems function and to interpret their results.17
4 Legislative and Judicial Responses
Legal systems are beginning to address Artificial Intelligence’s evidentiary impact through both legislation and case law. In South Africa, although no specific law governs AI-generated evidence, existing legislation, such as the Electronic Communications and Transactions Act (ECTA), recognizes electronic evidence as admissible if its authenticity and reliability can be proven.18 Internationally, courts in the United States and Europe have started to consider cases involving AI tools like facial recognition and algorithmic risk assessments.19 These cases emphasize the need for transparency and human oversight.
5 CONCLUSION
Therefore Artificial intelligence does play a crucial role in the legal framework of the law of evidence by making the evidence and information to be easily accessed. However it has some negative effects on the credibility and reliability of evidence, which in some case they defeat the end of justice in the legal system.
BIBLIOGRAPY
A) BOOKS
PJ SCHWIKKARD SE VAN DER MERWE, Principle of Evidence,p22 (2015)
B) CASE LAW
1) S v Zuma and Another (CCD30/2018) ZAKZPHC 39.
2) Mavundla v MEC: Department of Co-operative Government and Traditional Affairs KwaZulu-Natal and Other ZAKZPHC 2, 2025 (3) SA 534.
3) State v Loomis, 881 N.W.2d 749 (Wis. 2016).
C) LEGISLATION
Section 210 of the Criminal Procedure Act 51 of 1977
Electronic Communications and Transactions Act 25 of 2002.
D) JORNALS
- Ashley, K. D. Artificial Intelligence and Legal Analytics: New Tools for Law Practice in the Digital Age. Cambridge University Press(2017)
- Marwala Tshilidzi “AI And The Law Navigating The Future Together” United Nations University, UNU Center, 2024/02/13
- Goodman, B., & Flaxman, S. (2017). European Union regulations on algorithmic decision-making and a ‘right to explanation’. AI Magazine, 38(3), 50–57.
- Tania Broughton “Don’t rely on bots magistrate warns” 2023 July 17.
- Katz, D. M., Bommarito, M. J., & Blackman, J. (2014). Predicting the Behavior of the Supreme Court of the United States: A General Approach. PLoS ONE, 9(4).
1 PJ SCHWIKKARD SE VAN DER MERWE, Principle of Evidence,p22 (2015).
2Ibid.
3Ibid P23.
4Ibid p 23.
5 Section 210 of the Criminal Procedure Act 51 of 1977.
6 PJ SCHWIKKARD SE VAN DER MERWE, Principle of Evidence,p23 (2015).
7 S v Zuma and Another (CCD30/2018) ZAKZPHC 39.
8 Ashley, K. D. Artificial Intelligence and Legal Analytics: New Tools for Law Practice in the Digital Age. Cambridge University Press(2017).
9 Marwala Tshilidzi “AI And The Law Navigating The Future Together” United Nations University, UNU Center, 2024/02/13.
10 Ibid.
11 Ferguson, A. G. (2017). The Rise of Big Data Policing: Surveillance, Race, and the Future of Law Enforcement. NYU Press.
12 Tania Broughton “Don’t rely on bots magistrate warns” 2023 July 17.
13 Mavundla v MEC: Department of Co-operative Government and Traditional Affairs KwaZulu-Natal and Other ZAKZPHC 2, 2025 (3) SA 534.
14 Ibid.
15 Ibid.
16 Katz, D. M., Bommarito, M. J., & Blackman, J. (2014). Predicting the Behavior of the Supreme Court of the United States: A General Approach. PLoS ONE, 9(4).
17 Goodman, B., & Flaxman, S. (2017). European Union regulations on algorithmic decision-making and a ‘right to explanation’. AI Magazine, 38(3), 50–57.
18 Electronic Communications and Transactions Act 25 of 2002.
19 State v Loomis, 881 N.W.2d 749 (Wis. 2016).





