Lessons from the AI Government Readiness Index leaders

Randall Carolissen, Dean: Johannesburg Business School, University of Johannesburg. Photo: Supplied

Randall Carolissen, Dean: Johannesburg Business School, University of Johannesburg. Photo: Supplied

Published Apr 3, 2023

Share

Five years ago, geopolitical relationships were relatively stable, technology evolved at a manageable and predictable pace, and the economic growth of countries showed steady improvement following the 2008 financial crisis.

In 2023, record inflation levels are threatening global recession, simmering conflicts may spill over into full scale war in eastern Europe, and technology has rapidly accelerated - catalysed and precipitated by the Covid-19 pandemic.

Digital technologies are transforming markets and creating novel economic opportunities pervasively across all sectors. According to the World Bank, strengthening digital inclusion is of paramount importance for any nation’s competitiveness and social stability.

ChatGPT, launched towards the latter part of 2022, caused a step change in the impact that technology has in its potential to redefine all sectors, including education. It has had the fastest and most ubiquitous uptake of any of its innovation predecessors.

The World Bank estimates that 60% of global Gross Domestic Product in 2022 relied on digital communication technologies. This poses a significant risk to the post pandemic recovery of vulnerable populations unable to connect to, or use, digital technologies.

Research by Accenture has found that AI could double annual economic growth rates by 2035 by changing the nature of work, defining new synergies between man and machine, and boosting labour productivity by up to 40%.

The research compared the size of each country’s economy in a baseline scenario to expected growth in 2035 once the impact of AI has been absorbed into the economy. AI was found to yield the highest economic benefits for the US, translating to an additional $8.3 trillion (R147trl) in gross value added (GVA). In the UK, AI could add an additional $814bn to the economy by 2035.

Figure 1: Annual growth rates by 2035 of gross value added (a close approximation of GDP), comparing baseline growth by 2035 to an artificial intelligence scenario where AI has been absorbed into the economy. Source: Accenture

Yet profound obstacles to inclusion remain, and demographic divides continue to entrench inequalities across the global digital landscape. For example, one billion people, primarily in developing economies, are unable to prove their identity which limits their ability to access digital services and concomitant access to the levers of economic prosperity.

Furthermore, women are 7% and 16% less likely to own a mobile phone or use mobile internet, respectively. This means that 264 million fewer women than men have access to mobile internet. The share of internet users in urban areas is twice as high as in rural areas, and at the end of 2021, just 71% of the world’s younger population aged 15-24 was using the internet.

Africa still lags on several indicators essential for a successful digital revolution and if left unaddressed, will miss this window of opportunity for economic prosperity and to close historical and current digital divides. Africa must bridge the digital divide to leverage both its considerable youth dividend, as well as deploying technology for economic prosperity.

Figure 2: Africa’s ICT Indicators Source: Brookings

A well thought through sovereign AI strategy is imperative in enabling governments to improve both their responses to this increasingly dynamic landscape, and their obligations to their citizens. These rapid global shifts have caught most governments by surprise, and the less agile have found themselves struggling when it comes to appropriate regulatory responses.

This vacuum leaves space for the abuse and exploitation of technologically unsophisticated societies by less than salubrious agents and may even mitigate against efforts to roll back inequality and poverty, further polarising the globe and deepening divides. AI has proved to be effective in improving efficiency of service delivery, promote social justice and enhancing democracy.

The Oxford Insights AI Readiness Index, now in its fifth edition, offers governments the opportunity to benchmark themselves, and systematically identify gaps and shortcomings in their adoption of AI. It uses 39 indicators measured across 10 dimensions and grouped under three pillars: Government, Technology Sector and the Data and Infrastructure pillar.

In 2022 South Africa ranked 68 out of 181 countries surveyed globally, and second after Mauritius in Africa. The average score for sub-Saharan Africa of 29.4 out of 100 is considerably below the world average of 44.6, with 21 out of the 25 lowest scores achieved by countries in the region.

Predictably, the US is the index leader followed by Singapore. Analysing their scores may provide some vital clues for countries lower down on the totem pole.

Figure 3: Dimensions of AI readiness. Source: AI Government Readiness Index 2022

The North American region, consisting of the US, which ranked 1st and Canada which ranked 5th in 2022, constitutes the highest scoring region in the world with an average score of 81.6 against the world average of 44.6 out of 100. In both countries there is a strong appreciation of the potential benefits of AI in both the public and private sectors.

This translates into policy consistency even when there are political leadership changes, which scores high with investors and researchers as it engenders certainty for government investment in Research and Development (R&D), enhancing geopolitical trade in technology and protection of Intellectual Property (IP). Both countries are members of the Global Partnership on Artificial Intelligence (GPAI). The US also coordinates with European countries through the EU-US Trade and Technology Council, established in 2021.

The US’s performance on the technology sector pillar largely accounts for its top performance on the Index, with a score of 81.7, which by far outstrips its closest rival Singapore, with a score of 68.5.

According to the Index, the US’s high score in the maturity dimension indicates the dominance of later stage technology companies which offer off the shelf products for the government to procure. Similarly, a high score in innovation capacity indicates a robust start-up ecosystem to support innovation of novel AI applications which favours early-stage companies that can be price competitive on government contracts.

The US underperforms against its closest competitors on the Human Capital dimension. The Index notes that data privacy in the US is governed by a plethora of laws with varying rules for different types of personal data. The US National Artificial Intelligence Initiative (NAII) is a result of the National Defence Authorisation Act. With national defence as the dominant theme, NAII is meant to coordinate AI research, development, and education activities across US agencies and is supplemented by multiple committees and executive orders.

The Blueprint for an AI Bill of Rights, released by the White House Office of Science and Technology Policy in October 2022, outlines civil rights-related principles and provides a Technical Companion outlining how these principles can be applied while upholding democratic principles and fundamental rights. The bipartisan AI Training Act, which mandates that procurement officers inside federal agencies receive training on the advantages and disadvantages of AI, was also recently signed into law.

The dominant position of the US may cultivate anti-competitive behaviour by the biggest tech companies suppressing innovation in the long term, especially since they hold advantages in computing capacity and access to large datasets.

Big tech companies acquiring more innovative start-ups, investing in dedicated academic research centres and funding fellowship programmes could result in them shaping the overall academic research direction.

Singapore’s second position is owed to its performance on both the Government and Data, and Infrastructure Pillars, with scores of 89.7 and 94.2, respectively. Singapore performs very well in Digital Capacity and Adaptability dimensions due to its determination and effectiveness in implementing a dedicated and coordinated National AI strategy which was developed in 2019.

Singapore’s relatively lower scores on governance and ethics dimensions indicate more statutory work is required to protect citizens against harmful public sector innovation, and in obtaining more buy-in from the general populace. Singapore’s strength is derived from its innovative and commitment to digital government supplemented by a business-friendly legislative environment which leverages productive public-private sector partnerships to scaffold public sector innovation.

In the second part of this opinion editorial, we will look at Africa’s performance on the AI Readiness Index 2022, as many governments look to use AI to improve efficiencies in the area of service delivery – particularly in promoting citizens’ access to and experience of government services.

Randall Carolissen, Dean: Johannesburg Business School, University of Johannesburg, and Distinguished Professor (University of Woxsen).

BUSINESS REPORT