South Africa's AI Strategy: More Talk Than Action?

South Africa's AI ambitions face significant hurdles in implementation.

Article link: https://open.substack.com/pub/johanosteyn/p/south-africas-ai-strategy-more-talk?r=73gqa&utm_campaign=post&utm_medium=web&showWelcomeOnShare=true

Audio summary: https://youtu.be/k_JXEjfZdiA

I write about various issues of interest to me that I want to bring to the reader’s attention. While my main work is in Artificial Intelligence and technology, I also cover areas around politics, education, and the future of our children. This article delves into the critical intersection of AI, national policy, and its profound implications for South Africa’s societal and economic future, particularly concerning the next generation.

In July 2021, I penned an article for Business Day, “Hey Dad, 2050 Isn’t the Paradise You Imagined,” articulating my anxieties as a father raising a child in a world increasingly shaped by artificial intelligence. I argued then that the future my son would inherit hinged on our collective ability to govern technology responsibly. Four years on, my concerns have only deepened. South Africa, it seems, remains caught in a cycle of discussion, while the rest of the world actively forges ahead. My 11-year-old son, and millions of others like him, face a future potentially defined by unregulated systems that could entrench existing inequalities.

CONTEXT AND BACKGROUND

South Africa’s journey towards a national AI strategy began with considerable promise. The Department of Communications and Digital Technologies (DCDT) has been at the forefront, releasing a National AI Policy Framework in late 2024, intended as the first step towards a comprehensive National AI Policy or even an AI Act. This framework, which followed a discussion document in October 2023 and a National AI Summit in April 2024, aims to position South Africa as a leader in ethical and responsible AI, addressing national challenges and leveraging opportunities across the economy.

It outlines twelve strategic pillars, including talent development, digital infrastructure, research and innovation, ethical guidelines, data protection, and public sector implementation. The Western Cape government has even moved ahead with its own provincial AI policy for public healthcare, establishing ethical guardrails for innovation.

The initial working group convened by the National Advisory Council on Innovation (NACI), the Council for Scientific and Industrial Research (CSIR), the Human Sciences Research Council (HSRC), and the Department of Science and Innovation (DSI) was a beacon of what our country could be. It brought together diverse voices across genders, races, fields of study, and industries, united by a shared vision for an African-centred approach to AI governance.

Discussions revolved around protecting citizens without stifling innovation, ensuring data sovereignty to prevent digital colonialism, and fostering digital inclusion to leave no one behind. These principles are echoed in the National AI Policy Framework, which emphasises a human-centred approach, fairness, accountability, transparency, and mitigating bias.

INSIGHT AND ANALYSIS

Despite the robust framework and the initial collaborative spirit, tangible progress on the ground appears limited. While the policy development phase, including public consultations, has been completed, Cabinet approval, dedicated funding, and concrete implementation plans remain elusive. Minister Solly Malatsi acknowledged that policy development has been slower than anticipated due to the country’s multi-layered consultation and compliance processes. This inertia risks rendering the significant efforts of the working group a mere academic exercise. A policy without a budget is, as the saying goes, just a press release.

The framework’s high-level nature, without clear guidance on how ambitious goals will be implemented, further exacerbates this concern. This inaction stands in stark contrast to other African nations. Rwanda, Kenya, Egypt, and Mauritius have already established or funded national AI frameworks, moving from strategy to implementation. South Africa, despite its abundant talent and established institutions like the Centre for Artificial Intelligence Research (CAIR) and the AI Institute of South Africa (AIISA), is lagging.

This delay is particularly concerning given the rapid global advancements in AI and its transformative potential across sectors like healthcare, education, and public service delivery. Without a clear roadmap and allocated resources, South Africa risks falling behind, becoming a consumer rather than a creator in the global AI landscape. The implications for our country and, more personally, for the future of our children, are profound. Unregulated AI systems can exacerbate existing social and racial inequalities, influence critical decisions in employment and loans, and even infringe upon human rights.

The risk of algorithmic bias, where historical prejudices are perpetuated or amplified by AI trained on unrepresentative data, is a significant concern in a country with South Africa’s complex history. Furthermore, without local capacity to design, develop, and govern AI systems, we risk algorithmic colonialism, where foreign-developed AI models, shaped by external values, misinterpret local contexts and potentially silence legitimate voices. This directly impacts our children’s future, as they will navigate a world where critical services and opportunities are increasingly mediated by AI.

IMPLICATIONS

To bridge the gap between policy and practice, South Africa must take immediate, concrete steps. Firstly, the National AI Policy needs to be made public, approved by Cabinet, and adequately funded. This includes allocating resources for essential computing infrastructure in schools, developing open datasets, and expanding internet access, particularly in underserved areas. Secondly, the establishment of regulatory sandboxes is crucial.

These controlled environments would allow for the testing of real-world AI applications in vital sectors like education, health, and municipal services under independent supervision, fostering innovation while ensuring safety and ethical compliance. Thirdly, investing in AI literacy and skills development for our citizenry, especially our children, is paramount. They need to be equipped not only to use AI but also to critically question and understand its workings, ideally in their own languages.

Without these foundational steps, South Africa’s aspirations to leverage AI for economic growth, social equity, and improved public services will remain largely unfulfilled. The government must move beyond discussions and embrace decisive action to secure algorithmic sovereignty and ensure that the digital future serves all South Africans.

CLOSING TAKEAWAY

The promise of South Africa’s National AI Strategy is substantial, but its current trajectory suggests a concerning gap between ambition and execution. My initial worries as a father, four years ago, about the future my son would inherit, are still very much present. It is imperative that we move beyond “much ado about nothing” and translate our well-intentioned frameworks into tangible progress, ensuring a future where AI genuinely uplifts our nation and protects the prospects of our children.

Author Bio: Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net