South Africa’s artificial intelligence ambitions face a test beyond the conference floor: whether a country seeking to become a regional technology hub can establish credible rules for an industry whose own tools have already raised questions about the integrity of government policymaking.
AI Expo Africa returns to Johannesburg on October 28-29 for its ninth edition, bringing together technology companies, investors, government representatives and researchers to discuss responsible AI, sovereign computing, autonomous AI agents and the infrastructure needed to support adoption.
But the gathering comes after a controversy over South Africa’s draft national AI policy, which was withdrawn in June after allegations that parts of the document contained AI-generated material with fabricated or unverifiable sources. Parliament had called on the Department of Communications and Digital Technologies to explain the circumstances surrounding the document and the steps needed to produce a credible, consultative policy framework.
The episode exposed a central challenge for governments embracing generative AI: using the technology to accelerate policymaking is not the same as ensuring that policy is accurate, accountable or fit for implementation.
The distinction matters because AI policy is expected to establish safeguards for systems that can generate convincing but false information, reproduce biases and make consequential decisions with limited transparency. If the process for developing those safeguards fails basic standards of verification, the government risks undermining confidence in the framework before it takes effect.
The controversy does not establish that AI cannot assist with drafting legislation or policy. It raises a more fundamental question about who remains accountable for the output, how evidence is checked and whether officials can demonstrate that a policy reflects public consultation and independent expert judgment rather than unverified machine-generated material.
A growing market, but how deep is the domestic industry?
South Africa has a substantial commercial incentive to resolve those questions. MarketsandMarkets estimates that the country’s AI market will be worth $1.34 billion in 2026 and could expand to $6.77 billion by 2031, implying annual compound growth of 26.1%. The forecast reflects potential demand for AI software, services and infrastructure, rather than revenue earned exclusively by South African-owned AI companies.
Estimates vary considerably according to how researchers define the market. Grand View Research puts South Africa’s AI market revenue at $3.43 billion in 2025 and forecasts $4.92 billion in 2026. The difference between the estimates illustrates why market-size claims should be treated as forecasts based on different methodologies, not as interchangeable measures of the industry’s actual output.
The number of companies operating in the sector is similarly dependent on definitions. A September 2026 directory compiled by Auxo Digital tracks 39 notable South African AI companies, while a separate 2026 continental startup census cited by the directory counted 49 South African AI startups. Broader commercial databases classify many more organisations under AI-related industry categories, including companies that may use AI without developing their own technology.
Those figures suggest an ecosystem with room to expand, but they do not by themselves establish how much of the value generated by AI adoption will accrue to domestic developers. A market can grow rapidly while companies in the country remain dependent on foreign cloud providers, foundation models, computing hardware and intellectual property.
That distinction is particularly important for South Africa’s ambition to develop sovereign AI capabilities. The ability to deploy foreign-built models is not equivalent to having the infrastructure, research capacity, local datasets, skilled workforce and domestic companies required to build and maintain competitive systems.
The funding picture offers another measure of the challenge. According to Disrupt Africa’s African Tech Startups Funding Report 2025, South African AI startups raised about $22.8 million in disclosed funding during the year, accounting for 47.2% of the $48.3 million raised by AI startups across the six African markets covered in its reported breakdown. South Africa led that group, although the total remains small relative to the capital required to build large-scale computing infrastructure and develop advanced AI models.
The figures point to a policy challenge that extends beyond attracting technology vendors to Johannesburg. South Africa must create conditions in which local businesses can turn AI adoption into intellectual property, exportable products and sustainable employment, rather than primarily becoming customers of platforms developed elsewhere.
Responsible AI needs more than a conference
AI Expo Africa, backed by the South African AI Association, says its ninth edition will feature more than 50 speakers, industry panels, workshops and more than 100 AI brands. The programme is expected to cover enterprise deployment, cybersecurity, data centres, neocloud providers, energy-efficient computing and responsible AI.
Companies and organisations represented in the announced programme include Cassava Technologies, Clickatell, Huawei Technologies, Oracle, FirstRand Group, Sasol, the Financial Sector Conduct Authority and the National Institute for Theoretical and Computational Sciences.
These are consequential issues for a country trying to expand digital services while managing energy constraints, cybersecurity threats and questions about where sensitive data is stored and processed. More computing capacity can support AI development, but data centres also require reliable electricity, water, network connectivity and substantial investment. Their economic contribution depends partly on whether they stimulate local research, business development and employment.
The same scrutiny should apply to responsible AI. Corporate commitments to ethical deployment mean little without practical mechanisms for assessing model performance, documenting training data where appropriate, protecting personal information, auditing high-risk decisions and providing remedies when automated systems cause harm.
For government, the standard must be at least as demanding. Public-sector AI can affect access to services, employment, education, policing and the allocation of public resources. Officials therefore need to disclose where automated systems are used, establish clear lines of accountability and ensure that people can challenge consequential decisions.
The withdrawn policy also makes public consultation particularly important. A credible framework must reflect the country’s constitutional protections and social conditions, including unequal access to digital infrastructure and the risks of systems that perform poorly for some languages or communities.
Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.
Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.
Got a story to share? Contact Us to reach a global audience with Impact Newswire.
Mohd Hassan has extensive experience in news gathering, editing, and writing for the newswire industry, Contact – Info@impactnews-wire.com
Discover more from Impact Newswire
Subscribe to get the latest posts sent to your email.
