Sandile Hlophe | Partner | Leader | Africa Region Government & Public Sector | EY Africa | mail me |
Earlier this year, the South African government launched its National Artificial Intelligence (AI) Policy Framework. This is a strategic initiative aimed at harnessing AI for public value and socio-economic advancement.
The policy positions AI as a tool to enhance decision-making, improve public services and foster innovation across sectors. However, a recent survey we conducted with Oxford Economics, involving nearly 500 senior government executives across 14 countries, reveals a significant obstacle.
While governments globally recognise AI’s transformative potential, a substantial implementation gap hinders progress. This gap illustrates the disconnect between governments’ AI ambitions and reality.
Closing the gap between governments’ AI vision and reality
For South Africa, closing this gap is a strategic necessity. It must do so to meet rising citizen expectations and sustain public trust amid fiscal constraints and complex developmental needs.
The study underscores strong optimism among government leaders. Over 60% believe AI will deliver significant cost savings and enhance public services. A similar proportion anticipate that generative AI will revolutionise operations within five years.
South Africa’s AI framework reflects this ambition. It emphasises human-centred AI that supports evidence-based policymaking and promotes fairness in application. Yet, deployment lags significantly.
Only 26% of surveyed organisations have partially or fully implemented AI. Just 12% have adopted generative AI solutions. This implementation gap, where ambition outpaces action, demands immediate attention to prevent missed opportunities and reveals the ongoing tension between governments’ AI ambitions and reality.
Addressing the digital divide
South Africa’s policy framework, while forward-thinking, faces practical challenges that mirror global trends. It lacks robust integration with existing policies. This limits its coherence with broader governance structures. It offers insufficient strategies to address the digital divide, where data affordability and reliable connectivity remain significant barriers for many citizens.
Additionally, the framework does not fully explore AI’s potential to drive economic growth or its role in global supply chains. It also remains vague on funding mechanisms and implementation timelines. These shortcomings risk undermining the policy’s impact. They align with the survey’s warning that inaction could lead to operational inefficiencies and diminished public confidence.
The consequences of delay are considerable. Slow AI adoption could exacerbate service delivery gaps, strain fiscal resources and hinder South Africa’s ability to provide efficient, accessible services. In a nation with pressing socio-economic demands, hesitation may impede progress in critical areas such as healthcare, education and infrastructure. These are areas where AI could deliver transformative outcomes.
The role of public sector organisations in closing the gap
The survey highlights “pioneers”, public sector organisations that have successfully bridged this gap. For example, Victoria State’s Department of Transport and Planning in Australia uses an AI-powered data platform. This platform integrates transport data and provides predictive insights that improve management and commuter experiences.
Singapore’s Energy Market Authority employs AI to forecast energy demands. This strengthens infrastructure planning. These pioneers prioritise robust data systems, digitised processes and partnerships to address skill shortages. They demonstrate that a balanced approach to technology and organisational readiness is essential to close the divide between governments’ AI ambitions and reality.
South Africa’s own pioneer, the South African Revenue Service (SARS), exemplifies AI’s potential. SARS leverages AI-driven analytics to detect fraud and tax evasion. This enables precise, targeted audits. Predictive models assess compliance risks and optimise resource allocation. Automation of routine tasks, such as tax return processing, allows staff to focus on complex issues.
Plans to integrate AI with other government systems will enable SARS to enhance data sharing and enforcement. This aligns with global best practices. These AI initiatives strengthen fiscal stability, which is critical for funding social programs. SARS illustrates South Africa’s capacity to lead in AI adoption. However, scaling such efforts across other sectors remains a pressing priority.
Turning AI ambitions into action
The survey identifies several barriers that South Africa must address. These include privacy and security concerns, misaligned strategies, inadequate infrastructure and ethical considerations. Pioneers with implementation experience recognise these challenges more acutely. This suggests that complexity emerges through practical engagement.
South Africa’s digital divide and shortage of AI-skilled professionals underscore the need for targeted interventions. To address these challenges, South Africa can adopt a multifaceted approach.
Encouraging public-private partnerships can leverage private sector and academic expertise. This will accelerate AI innovation. Investing in AI-focused education from primary to tertiary levels will equip the workforce for the digital economy. Supporting local AI startups through grants and mentorship can foster innovation and job creation. This can be amplified through private sector collaboration.
Collaborating with global organisations
Pilot projects in healthcare, agriculture and education can demonstrate AI’s benefits. These pilots will build momentum for broader adoption. Additionally, engaging in international collaboration with global organisations can enhance capabilities through shared knowledge and best practices.
By addressing privacy, security and ethical concerns, these strategies can create a robust AI framework. This framework can drive innovation, economic growth and social development.
Public trust is critical to this effort. The study notes that nearly half of citizens across 15 countries have recently used AI-driven government services. This indicates growing acceptance. However, trust depends on transparent data practices and ethical frameworks that mitigate bias. South Africa must prioritise these elements. Doing so will ensure AI serves the public interest while addressing privacy concerns and fostering confidence in AI-driven services.
Globally, the 14 countries surveyed, including Australia, Canada and the United States, face similar pressures. These include fiscal constraints, workforce shortages and geopolitical complexities. AI offers a pathway to address these issues. However, success depends on decisive leadership.
In conclusion
South Africa’s pioneers, like SARS, demonstrate that progress is achievable. AI could optimise healthcare delivery in underserved areas. It could also predict agricultural yields to support farmers or enhance access to quality education. These goals align with proven use cases, such as AI chatbots for citizen engagement and predictive analytics for resource allocation.
The study calls for immediate action. For South Africa, the implementation gap is both a challenge and an opportunity. Delaying action risks undermining the government’s ability to fulfil its responsibilities.
A strategic approach, built on data foundations, partnerships and a commitment to trust, can unlock AI’s potential. This potential can improve services, boost efficiency and increase public confidence.
The nation’s future depends on decisions made today.
