The implications of POPIA in the use of AI in South African litigation

Jul 25, 2024 | News

The Protection of Personal Information Act, 4 of 2013 (POPIA) is a crucial piece of legislation in South Africa that governs the utilisation of personal information.  It further protects and guides the usage, processing, storing of personal information and aims to protect individuals’ privacy rights.

When it comes to the use of Artificial Intelligence (AI) in South African litigation, POPIA has several implications that legal practitioners and organisations need to consider when making use of AI in litigation proceedings.

Here are some of the key implications of POPIA on AI in South African litigation:

Data Collection: AI tools have the capability to collect voluminous amounts of information. It is imperative for legal practitioners to verify the sources of the data in which they input into AI tools, as well as the generated information which is received from AI tools. Doing so ensures that AI tools avoid generating information that is fictitious, and subject to AI bias.

Data Processing: AI tools have the capability to assist legal practitioners in processing high volumes of information such including the following:

  • Extracting pertinent details: This includes reviewing documents, case law and agreements to pinpoint critical information such as involved parties, previously referred to decisions, dates, obligations, and conditions.
  • Comparing terms: This task entails scrutinising analogous clauses across diverse documents to observe parallels, disparities, and inconsistencies.
  • Flagging anomalies: Anomalies may comprise of discrepancies in wording, contradictory terms, or unconventional conditions.

In light of the above, under POPIA, legal practitioners must also ensure that they have the necessary consent from their clients to process their personal information. Fee mandates should include clauses dealing with their consent in respect of data collection and processing.  Legal practitioners’ clients, including their intermediaries, agencies and bureaus, should ensure that that any information provided to their legal representatives, which may form the basis for potential litigation, is collected in an ethical and transparent manner with the necessary consent where required.

Data Minimisation and Purpose Limitation: POPIA requires that organisations limit the collection and processing of personal information to what is necessary for a specific lawful purpose. When implementing AI in litigation, legal practitioners need to ensure that AI systems only access and analyse personal data that is relevant to the legal case at hand. Data minimisation practices should be followed to mitigate the risk of unauthorised access or use of personal information. When making use of AI powered tools to sift through dense documentation or locate a legal precedent relevant to a case on hand, litigators should be cautious when uploading sensitive information which may be legally privileged.

Data Security and Integrity: POPIA mandates that organisations implement appropriate security measures to safeguard personal data from unauthorised access, disclosure, or destruction. When using AI in practice, legal practitioners must ensure that the AI systems or tools be used adhere to robust data security standards and POPIA policies to protect sensitive information belonging to their clients.

Accountability and Compliance: POPIA places a strong emphasis on accountability and compliance with data protection principles. Legal practitioners utilising AI in litigation are responsible for ensuring that personal data is processed lawfully, transparently, and securely. Organisations must maintain detailed records of AI processes and data flows to demonstrate compliance with POPIA regulations. Data protection impact assessments and reports should be conducted and implemented to identify and mitigate privacy risks associated with AI implementations in litigation.

Cross-Border Data Transfers: POPIA regulates the transfer of personal data outside of South Africa to ensure that adequate levels of data protection are maintained. When utilising AI technologies that involve cross-border data transfers in litigation, such as the sharing and processing of client information outside the boarders of South Africa, legal practitioners must ensure that appropriate safeguards are in place to protect personal information. For example – when dealing with matters involving parties in multiple jurisdictions, South African practitioners must ensure that foreign practitioners adhere to, inter alia, binding polices, client consent and lawful purposes that ensure the protection of the information received from South Africa.

Whilst the advantages of AI in litigation are evident, one must be cautious about the unethical or misuse of AI. In a recent decision by a South African Court, a client’s legal practitioner relied solely on the use of AI to generate legal research, which they relied upon and placed before the court. The information provided to the court was false and inaccurate, as a result of being fabricated by AI. Not only was this highly unethical but led to the detriment of their client’s case. Legal practitioners are cautioned against being complacent, and relying solely on the use of AI to assist them in their client’s legal needs. When utilising AI in legal proceedings, thorough consideration and analysis of generated information is imperative to ensure compliance with the provisions and protection afforded by POPIA.

By Jodi Poswelletski | Director and Jason Berkowitz | Associate

Discuss this article with me

Recent Articles

When Can a Medical Scheme Cancel Membership for Non-Disclosure?

When Can a Medical Scheme Cancel Membership for Non-Disclosure?

11 Things I Wish I’d Known Before Starting My Articles

11 Things I Wish I’d Known Before Starting My Articles

What’s in a Drop? The Case of Cape Velvet and the Price of Classification

What’s in a Drop? The Case of Cape Velvet and the Price of Classification

REPORT: AI in the Rainbow Nation: Shaping the Future of AI Governance through Global Insights

REPORT: AI in the Rainbow Nation: Shaping the Future of AI Governance through Global Insights