THE IMPACT OF ARTIFICIAL INTELLIGENCE ON EVIDENCE LAW

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).