Nearly eight months after unveiling its National Artificial Intelligence (AI) Policy Framework, South Africa is finally opening the document for public review this April. The Department of Communications and Digital Technologies (DCDT) will spearhead the review process, inviting feedback on the framework, which aims to guide ethical, inclusive, and innovation-driven AI development across the country.
While the framework touches on vital issues such as talent development, data protection, and responsible innovation, industry leaders have voiced frustration over its sluggish progress.
“There’s a great deal of uncertainty,” said Nerushka Bowan, a technology and privacy lawyer and founder of the LITT Institute. “What South Africa urgently needs is a clear, actionable roadmap that sets out the direction and regulatory approach for AI.”
Despite its status as a tech leader on the continent, South Africa is trailing behind countries like Rwanda, which approved its national AI policy in April 2023. Rwanda’s framework focuses on using AI to drive economic growth, improve public service delivery, and foster ethical AI adoption. Experts warn that South Africa’s delay could cost it a share of the estimated $1.5 trillion that AI could add to Africa’s GDP by 2030.
Daniel Novitzkas, Group Director at digital solutions firm Specno, emphasized the strategic opportunity being missed. “Without a defined AI policy, South Africa risks losing competitive ground, not just economically, but in attracting investment and talent.”
Legal experts have also stressed the importance of finalizing the policy. Wendy Rosenberg, Director and Head of Digital Media and Electronic Communications Practice at Werksmans Attorneys, acknowledged that while the framework addresses critical themes like privacy, governance, and transparency, the delay is hampering implementation.
“Establishing this framework is essential—it lays the groundwork for industry-specific policies that can regulate AI across sectors such as finance, healthcare, and education,” Rosenberg noted.
Currently, AI is being governed piecemeal through existing sectoral laws, leaving developers and investors uncertain about future legal requirements. Bowan warned that this regulatory vacuum could discourage foreign investment and push local talent to countries with more defined AI regulations, such as the U.S., U.K., and Canada.
“Investors are waiting for clarity. Without it, we risk losing the very innovators capable of transforming our economy,” Bowan said.
While South Africa benefits from existing laws like the Protection of Personal Information Act (POPIA)—which aligns with the EU’s GDPR—experts argue that AI regulation must go further. Rosenberg stressed the importance of incorporating ethics, transparency, and bias mitigation into AI governance.
Given the country’s socio-economic inequalities, ensuring fairness in AI deployment is especially important. Rosenberg called for mechanisms such as human-in-the-loop systems, diverse data sets, and ongoing bias evaluations to help ensure accountability and prevent discriminatory outcomes.
Beyond policy direction, infrastructure gaps such as internet access pose another hurdle. Approximately 28% of South Africans were still offline in 2023, limiting the inclusive potential of AI technologies.
“Even with powerful AI tools available, large segments of our population lack the digital access needed to benefit from them,” Novitzkas said.
Experts advocate for a risk-based approach, similar to the European Union’s AI Act, which imposes stricter rules on high-risk applications while allowing flexibility for low-risk use cases. Such an approach, coupled with future-ready and principle-based legislation, could help South Africa leapfrog into a leading position in AI development.
“Regulation should guide innovation, not hinder it,” Rosenberg concluded. “If South Africa can strike the right balance, AI could be a game-changer for investment, job creation, and digital transformation.”
