South Africa’s Department of Communications and Digital Technologies has published a draft national artificial intelligence policy for public comment, built around six strategic pillars and a phased three-year implementation plan, with full implementation expected in the 2027-2028 financial year.
“The aim of the AI policy is to ensure that both the benefits and risks brought by AI are evenly distributed across society and generations,” Cabinet said in a statement approving the policy’s publication.
The draft is based on the South African National Artificial Intelligence Policy Framework published in August 2024 and was informed by 32 stakeholder submissions as well as consultations through the Cabinet Cluster process. The department described it as a work in progress requiring extensive external consultation before finalization.
The policy’s six strategic pillars are capacity and talent development; AI for inclusive growth and job creation; responsible governance; ethical and inclusive AI; cultural preservation and international integration; and human-centered deployment.
On talent development, the department said integrating AI into school curricula from primary through tertiary level will be essential, including training educators in AI tools and incorporating multilingual education where possible. On inclusive growth, the country plans to establish and sponsor specialized AI research institutions focused on building scientific knowledge in machine intelligence. The safety and security pillar includes installing robust controls to safeguard AI systems and establishing legal and technical remedies for deepfakes and defamatory AI content.
The ethical and inclusive AI pillar focuses on fairness, transparency, accountability, inclusivity, confidentiality and human autonomy, with the department emphasizing that preserving democracy is a key concern. “The preservation of democracy is a key focus of AI development and deployment to ensure that AI systems support democratic values rather than undermine them,” it said. The cultural preservation pillar seeks to maintain South African cultural values while enabling global integration, while the human-centered deployment pillar focuses on maintaining human control in AI decision-making processes, particularly in generative AI.
A concept central to the policy is intergenerational equity — the need to ensure AI development serves both current and future generations. “This involves embedding sustainability and ethical foresight into AI strategies to create long-term value for society while mitigating risks,” the department said.
Implementation is structured in three phases, each covering one financial year. The first phase, covering 2025-2026 and now complete, involved finalizing the draft policy, identifying key regulatory requirements and initiating the development of national AI policy guidelines. The second phase, running through March 2027, will cover publishing national AI policy guidelines, implementing regulatory requirements for high-risk AI use cases, identifying draft requirements for medium- and low-risk use cases, developing sectoral AI strategies and beginning the institutional framework design. The third and final phase, set for 2027-2028, will see full implementation of outstanding policy interventions, with provision for updates as AI technologies evolve.
A significant structural decision is the choice not to create a single AI regulator. Instead, oversight will be distributed among existing regulatory authorities — including the Information Regulator for privacy protection — in a coordinated, multi-regulator model. Law firm Baker McKenzie described this as a decision that “represents a coordinated oversight approach rather than the creation of a centralised AI regulator.”
In the interim, AI-related activities remain governed by existing legislation including the Protection of Personal Information Act, the Consumer Protection Act, the Electronic Communications Act, the Electronic Communications and Transactions Act and the Cybercrimes Act. Legal practitioners say the new policy is expected to lay the groundwork for a future National AI Act.
South Africa’s AI policy process began with the release of a draft national AI plan discussion document in April 2024. Some legal and technology observers have expressed concern about the pace of development, with policy finalization now expected only in 2027.
