South Africa’s public sector has much to gain from AI, but outdated policies, data security concerns, limited resources and fear of unintended consequences are making officials wary of embracing the technology.
Smart applications such as artificial intelligence (AI) and automation are touted as the silver bullet that can assist the public sector to significantly improve service delivery in South Africa by automating routine tasks, streamlining workflows and enhancing citizen engagement.
The efficient delivery of basic services to the electorate seems to be a goal that eludes many municipalities who are at the coal face of interacting with communities. Successive reports by the Auditor-General (AG) makes for very grim reading and paint a picture of widespread malfeasance and ineptness which hampers the ability of the state to provide basic services to its electorate.
The AG’s reports point to systemic institutional decay, poor financial control and a severe lack of accountability across South African municipalities, with factors such as negligent, incompetent, corruption, or malicious actions by public officials, instability and a lack of qualified staff in senior management, are being cited as some of the major reasons for this state of affairs.
However, if the adoption of smart applications is a panacea that can address service delivery challenges, then why are public sector entities and all spheres of government reluctant to integrate these solutions in their operational processes?
The answer to this is not simple and clear cut. Firstly, we have an outdated policy framework such as the Public Finance Management Act (PFMA) which restricts innovation and development in the public sector. While the act has great intentions, it is laden with bureaucratic hurdles that considerably slows down delivery.
Our current policy framework is archaic and has not kept abreast of the rapid technological changes that have swept across many sectors of the economy and society. Our budgeting orientation is backward focused rather than forward looking because we work on what we have been given. If we want to employ a tool like AI, which is descriptive focused, it becomes a challenge because it presupposes a budgetary reform which will not materialise because of the budgetary policies that are in place.
Secondly, there is a lot of apprehension in the public sector about the potential compromise of the security integrity of sensitive information that may be precipitated by the adoption of AI. Many senior officials fear that going the AI route might have unintended consequences of opening the back door to cybercriminals who may access classified data that may compromise national security.
The public sector does not have sufficient skilled personnel, the budget nor the infrastructure required to safeguard this data, which then necessitates the outsourcing of this function to external service providers whose services are prohibitively expensive to secure.
Thirdly, the public has a mandate of delivering services to the public. Given a choice between investing in new technologies that show no immediate political and service delivery outcomes on the one hand and funding tangible projects that deliver immediate and visible results, officials in the public sector readily opt for the latter option.
Senior public sector officials and finance officials regularly weigh the options available to them and channel the available resources to public priorities such as patching the potholes, or fixing the water and electricity infrastructure for example, where the results are immediate, impactful and visible.
This largely explains why the investments in technology advancements by the public sector are minimal. Having a functional health, judicial, or education system or a good road network takes priority over investments in smart applications.
Unlike the private sector where the shareholder characterises a return on investment in terms of dividends, our stakeholders who are tax and ratepayers, define value return on the quality of service we deliver to communities. Our shareholders who are our political principals, demand an investment in priority areas that gratifies their electorate.
In a country with a stubbornly high unemployment rate, the state is less inclined to invest in technologies that have a potential to shed jobs. The retention of jobs is very critical in the public sector hence the emphasis is using the available tools to improve operational efficiencies.
The fear of change, coupled with an environment that is less agile and receptive to doing things differently, serves as the fertile ground for resistance for any innovation. Unlike in the private sector where innovation is encouraged and rewarded, there is a low margin for error in the public sector as any financial losses that can be incurred from such investments may be deemed as fruitless and wasteful expenditure that triggers an enquiry and an investigation.
As a result, financial professionals in the public sector are wary of investing in smart solutions because of the real and present danger of a potential fallout that may permanently sully their professional integrity and subject them to public and media scrutiny.
Smart technologies have an immense potential to flip the script and change the service landscape across all spheres of government, be it at national, provincial and municipal level. However, the structural, policy and human frailty are stifling any attempts to leverage the power of technology to improve operational efficiencies.
But all is not doom and gloom. Encouragingly, there are discussions in the corridors of public sector entities on how best to integrate into the public sector to enhance governance and productivity. Nonetheless, AI discussions at this stage should not look into potential integration of these solutions, but rather the dialogue should focus on building a framework that will allow the implementation of AI in the public sector. That should be the point of departure.
