South Africa has taken a major step toward regulating artificial intelligence after the Department of Communications and Digital Technologies (DCDT) released a draft national AI policy designed to guide future laws, industry standards and government oversight.
The proposed framework introduces a phased regulatory roadmap built around six strategic pillars aimed at supporting innovation while addressing risks linked to privacy, misinformation, deepfakes and job disruption.
Cabinet recently approved the draft for public comment, marking the latest development in the country’s digital policy agenda.
Rather than creating a single AI watchdog, the government is proposing a sector-based model in which oversight responsibilities are distributed across existing regulatory bodies.
This includes entities such as the Information Regulator, which is expected to play a key role in data privacy and information governance.
Officials said the policy is based on the National Artificial Intelligence Policy Framework first published in August 2024 and refined through stakeholder consultations and Cabinet cluster processes.
Three-year implementation roadmap
The rollout has been divided into three financial-year phases.
The first phase, completed in the 2025/26 financial year, focused on finalising the draft policy, identifying urgent regulatory priorities and beginning the development of national AI guidelines.
The second phase, now underway and running until 31 March 2027, will focus on publishing policy guidelines, introducing rules for high-risk AI use cases and drafting regulatory requirements for medium- and low-risk systems.
Sector-specific AI strategies and institutional funding models are also expected to be developed during this period.
The final implementation stage is planned for the 2027/28 financial year, when remaining policy interventions are expected to be fully operational.
Six strategic pillars define the policy
The draft policy is centred on six core pillars:
- capacity and talent development
- AI for inclusive growth and job creation
- responsible governance and security
- ethical and inclusive AI
- cultural preservation and international integration
- human-centred deployment
A strong focus has been placed on education and skills development, including the integration of AI into school and tertiary curricula.
Government also wants to support specialised research institutions and local machine intelligence development.
Another key pillar is the legal and technical response to AI-generated misinformation, including deepfakes and defamatory content.
The policy further stresses democratic safeguards, transparency and human oversight, especially in generative AI systems.
Officials said the strategy seeks to balance technological progress with constitutional values and long-term social benefit.
Source: Department of Communications and Digital Technologies / South African Cabinet statement
