Artificial intelligence is increasingly influencing how South Africans live, work, and access services. Until recently, this evolution has largely occurred without a formal national framework.
The South African government has released a draft national AI policy for public review, initiating a process to establish guidelines for artificial intelligence governance within the country. Approved by the Cabinet on March 25, the draft policy is open for public input until June 10.
Built on the Constitution and Public Interest
The draft policy is guided by South Africa’s Constitution. The Department of Communications and Digital Technologies, responsible for the policy’s development, states that the framework is rooted in constitutional values, human rights protections, and the pursuit of inclusive economic growth. While AI is currently utilized in both government and private sectors, it operates under fragmented internal rules rather than a unified national law. This draft policy aims to establish a more structured national approach.
Not Yet Law, But Already Influential
Legal experts at Bowmans law firm consider the draft an important initial step in shaping South Africa’s AI landscape. Robby Coelho, Justin Ramages, Ashleigh Brink, and Tanya Chivauras note that the policy is not currently legally binding but signals the direction of future regulation. The document serves as a starting point, outlining the government’s intended approach to AI governance and already influencing expectations regarding ethics, accountability, and transparency.
AI as a Tool for Growth and Inclusion
The draft policy positions AI as a catalyst for economic inclusion, job creation, cost reduction, and development across Africa. It identifies key sectors where AI could have a significant impact, including education, healthcare, agriculture, and public administration. Simultaneously, the policy emphasizes addressing South Africa’s inequality gap, digital divide, and broader socioeconomic challenges. AI systems must be developed and deployed in alignment with human rights standards and the public interest.
New Systems for Oversight and Accountability
A significant element of the draft is its proposed institutional structure. Plans include the creation of a National AI Commission or Office, an AI Ethics Board, an AI Regulatory Authority, an AI Ombud’s Office, a National AI Safety Institute, and an AI Insurance Superfund. These institutions would form the foundation of future AI oversight in the country. The policy also proposes a phased regulatory approach, beginning with high-risk AI applications and gradually expanding to medium- and low-risk systems.
A Phased and Risk-Based Approach
Rather than implementing broad regulations, the draft suggests a risk-based model. This means strict regulation for “unacceptable-risk” AI uses, targeted rules for high-risk systems, and gradual guidance for medium- and low-risk applications. Sector-specific working groups and implementation roadmaps are also included in the plan. This approach reflects global trends in AI governance, where regulation is increasingly tailored to risk levels.
What it Means for Businesses and the Public
Although the policy remains in draft form, legal experts indicate that its publication already impacts companies working with AI. It signals a stronger regulatory focus on AI ethics, transparency, data protection, fairness, and sector-specific oversight. Businesses in South Africa’s digital economy are expected to prepare for a future with more structured and enforceable AI governance.
Public Participation Now Becomes Key
With the policy open for public comment until June 10, South Africans have an opportunity to provide input. This stage is crucial for government, industry, and everyday users whose lives are increasingly shaped by AI systems. The outcome of this process will likely determine how South Africa balances innovation with rights, regulation, and inclusion in the age of artificial intelligence.
