government policy Artificial intelligence in South Africa will be finalized in the 2026/2027 financial year, with the government expected to soon launch an implementation plan following a recent symposium on the development and implementation of an AI policy framework.
The draft policy will be discussed by the Economic Cluster Ministerial Council next month, before going to the Cabinet committee. It is expected to be gazetted in late March for public comment for 60 days.
referring to the frameworkDepartment of Communications Deputy Director-General Alfred Mamoto, who heads policy development and strategy, told Parliament's portfolio committee on communications and digital technologies that between now and June, a quarterly digital economy stakeholder forum will be set up with five workstreams.
The Ministry of Communications will establish a priority working group involving the department, the South African Local Government Association and the Department of Cooperative Governance to improve the implementation of the bylaws and launch a concept document for a “National Evidence Centre”.
From July to December next year:
- A regulatory forum will be operated;
- The National Digital Skills Framework will be reviewed and updated;
- The Department will establish and resource an investment desk and start-up desk, launch an ICT sector regulatory sandbox and develop a legislative coordination road map; And
- Establish and maintain South Africa's position in global digital platforms.
Stakeholders will gather for a regulatory impact assessment on quick wins and begin a process to define what meaningful access means.
South Africa began to develop its own policy Aye In 2020. Although it has improved its consultation process, policy development has been slower than some of its international peers. At present the country is relying on self-regulation.
On Tuesday, MPs were told what the purpose of the policy was.
Responsible AI
Mamoto said the policy is based on 14 pillars, including education, training, industry collaboration, digital infrastructure, innovation, ethical guidelines, security and privacy.
Although the policy will be comprehensive, sector-specific rules will have to be compiled, Mamoto told the committee.
On capability and internal development, it will focus on building national AI skills inventories through education, training and industry operations. These have to be enabled by strong digital infrastructure and connectivity.
Reading: How AI is rewriting the rules of consulting
The policy will make proposals on responsible AI governance, including addressing challenges related to safety, security and privacy.
“I think I can emphasize this a little bit… We've seen a lot of disinformation and disinformation that is done using AI. So, we need to make sure that we have some guardrails to make sure that we don't perpetuate this and we also don't spread, you know, these deepfakes,” Mamoto said.
“And one of the areas that this highlights is that you will find that AI systems that are not trained by local data sets are the most likely to have bias. So, one of the things we advocate in the policy is that whenever you are trying to train an AI system, you should have data sets that are inclusive and that take into account all the demographics of the country.”
The document also discusses accountability in developing AI systems and “how can we employ you if the system you develop could harm society”, he said.
It advocates cultural preservation and global integration, as well as strengthening international cooperation and competitiveness.
“On top of this, what we are looking at doing is to ensure that we can develop language models, especially for those languages which have less speakers.
“But at the same time, we would like to use AI to digitize some of our indigenous knowledge systems as well as music and art to ensure that it can be used for economic development…” Mamoto said.
The main focus of the policy is to ensure that there is always a human being at the center of AI development.
He said, “We need human oversight, especially if we have to see, let's say you apply for a particular service in the government. You know, we can't just use AI as a black box….”
To support innovation, “regulatory sandboxes” will be set up in controlled environments to test AI systems and look at associated risks.
Midway
Some of the discussion on Tuesday focused on the benefits of regulating AI.
Acting committee chairman Sheikh Imran Subrathy said that for example, in India he believes over-regulation stifles innovation, while in the EU it is the opposite. Last year, the Free Market Foundation warned against South Africa is blindly following EU rules on technology and AI.
Mamoto said that after a “benchmarking exercise”, the department agrees that the EU rules are of concern.
Reading: African companies are working on cloud and AI – at least on paper
He said, “Our path in South Africa is the middle of the road. We have to adopt this policy to ensure that we have a policy lever on which we can stimulate economic growth, ensure social well-being, but also ensure that we deliberately position our country as a leader in innovation.” — (c) 2026 NewsCentral Media
Get breaking news from TechCentral on WhatsApp. Sign up here.