Digital evidence

Defending against AI-enhanced surveillance

In today’s tech-driven world, artificial intelligence (AI) and 5G technology have significantly influenced surveillance, law enforcement and evidence collection. AI-powered tools include facial recognition, predictive analytics and automated licence plate readers, which are increasingly used in criminal and civil cases. However, despite their reputation for accuracy, these technologies are not fool-proof. South African courts must carefully evaluate AI-generated evidence, ensuring its reliability and fairness before admitting it. 

South Africa’s legal framework for digital evidence

South African law recognises digital evidence under the Electronic Communications and Transactions Act 25 of 2002 (ECTA), which sets the standard for admissibility. The Law of Evidence Amendment Act 45 of 1988 further allows electronic records as evidence, provided their authenticity and reliability are proven. The Cybercrimes Act, which took effect in May 2021, outlines cybercrimes and establishes guidelines for investigating, searching and seizing digital evidence. However, as AI-powered surveillance and 5G technology become more prevalent, new legal challenges arise, particularly regarding accuracy, bias and potential violations of constitutional rights under the Protection of Personal Information Act (POPIA).

Key concerns with AI-generated surveillance evidence

AI surveillance systems are not without flaws. Facial recognition and pattern analysis can incorrectly identify individuals, especially those from diverse racial or ethnic backgrounds, leading to false accusations. AI models should be interrogated for accuracy and transparency of algorithms, because AI systems often reflect biases in training data, unfairly targeting certain groups. If evidence is based on imperfect AI models, its reliability is compromised. Have independent assessments validated its fairness? If not, it should not be relied upon.

Integrity of digital evidence is vital to prevent tampering or misinterpretation. AI-powered surveillance tools produce vast amounts of data, which can be altered or misused. The chain of custody should be clear and well documented to confirm the authenticity of digital evidence in court.

AI-driven surveillance and data collection raise concerns about privacy. particularly under Section 14 of the Constitution. If evidence is obtained through intrusive or unlawful means, it can be challenged in court as a violation of constitutional rights. 

AI can be used to generate convincing deepfake videos and synthetic voice recordings, posing a risk to digital evidence integrity. False video or audio recordings can potentially lead to wrongful convictions or misleading legal proceedings. As these technologies improve, the risk of manipulated evidence increases.

Legal strategies for challenging AI-based digital evidence

If an attorney is faced with digital evidence in a case, there are ways to ensure the evidence is safe. They can request details of the system’s source code, training data, error rates and bias testing to assess its credibility. Digital forensic experts can challenge digital evidence by identifying flaws, biases or improper usage.

If a lawyer suspects AI surveillance of having been unlawfully used, they can argue for the exclusion of evidence under Section 35 of the Constitution, which protects the right to a fair trial. Lastly, the chain of custody should be evaluated. This includes determining how the digital evidence was collected, stored and handled. This may reveal weaknesses which could render the evidence inadmissible.

Strengthening the legal system’s approach to AI-generated evidence

The Cybercrimes Act strengthens existing laws by ensuring digital evidence is handled in a way that balances privacy with authenticity. Forensic techniques can detect manipulated media, including digital alterations. Our legal system must develop regulatory frameworks for AI-generated evidence and provide specialised training for judges, attorneys and law enforcement officers on AI’s strengths and limitations before courts can safely rely on digital evidence.

The future of AI, 5G and legal defence

As AI and 5G technology continues to advance, legal professionals need to be vigilant and review their strategies for dealing with digital evidence. Expertise related to digital forensics, data privacy laws, AI ethics and cybercrime regulations is fast becoming essential for all law firms, whatever their area of practice. Courts should avoid accepting AI-generated evidence without thorough examination to protect the rights of the accused and uphold justice.

For modern litigators, an understanding of AI’s role in evidence is no longer optional. Attorneys who put their clients first will actively question the reliability and fairness of AI-generated evidence to ensure they safeguard their clients’ rights.

SD Law can help

At SD Law, we welcome tools that enhance our service to our clients and the advancement of law enforcement and the legal profession in general. We embrace cutting-edge technology as part of our vision to be a modern, client-driven law firm, and we stay informed about the latest advances in the use of AI in criminal and civil law. If you think you have been a victim of discrimination or unfair proceedings due to AI tools in law enforcement or unreliable digital evidence, we want to hear from you. We can help. Contact Simon on 086 099 5146 or email  sdippenaar@sdlaw.co.za for a confidential discussion.

Further reading:

• AI in litigation
• AI in law enforcement

SD Law

View more posts from this author