The Department of Communications and Digital Technologies has published a draft national policy on artificial intelligence (AI) for South Africa. The department said the plan rests on six main pillars meant to grow the local AI sector.
According to the draft, AI rules would be handled sector by sector, with several regulators sharing the work. The department wants AI oversight built into existing bodies, such as the Information Regulator, to help protect people’s privacy.
“The Draft National AI Policy is based on the South African National Artificial Intelligence Policy Framework of August 2024,” the department said in a notice in the Government Gazette.
The department said it drew up the policy using 32 submissions from interested groups on that framework. It also consulted government structures through the Cabinet Cluster process.
The department said the draft is still a work in progress. It said the final plan will need wide consultation outside government.
“Accordingly, the department envisions a staged implementation approach to enable government to respond to the most pressing regulatory needs immediately,” it said.
The rollout is split into three stages, each lasting one financial year. According to the department, the first stage ended on 31 March 2026. It covered finishing the draft policy, publishing key rules and starting work on national AI guidelines.
The second stage runs in the current financial year, which ends on 31 March 2027. The final stage is set for the 2027/28 financial year. The department said the remaining steps may need updates as AI technology changes.
The first pillar deals with skills and training. The department said AI should be part of school and university lessons, from primary level up to tertiary education.
“This entails the education of educators in AI tools and the integration of multilingual education where possible,” the department said.
The second pillar links AI to growth and jobs. Under this pillar, the department said the country plans to set up and fund special AI research institutions.
The third pillar covers safety and security for citizens and infrastructure. The department said South Africa must create strong technical and legal remedies against deepfakes and defamatory AI content.
The fourth pillar focuses on ethical AI. The department listed fairness, transparency, accountability, confidentiality and human autonomy among its principles.
“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,” the department said.
Another pillar aims to protect South African culture while allowing the country to work with other nations on AI.
The last pillar calls for people to stay in control of decisions made by AI systems, especially generative AI. The department said the strategy uses human feedback to train AI systems so they follow ethical rules and social norms.
Source: this article is based on reporting by MyBroadband. Image: MyBroadband. Written with the help of AI and published by the Tzaneen Voice Technology desk. See our Editorial Standards.
Werner Jacobs covers technology for Tzaneen Voice. Werner reports on mobile networks, data prices, internet access, smartphones, apps, AI, cybersecurity and South African tech companies. Each story explains what a new product, price change or policy means for ordinary users in South Africa.