Article Preview
TopIntroduction
In the last decade, significant improvements in connectivity, processing power, and cheap data storage like the cloud have led to the focus on huge volumes of data generated in both the public and private sectors (Merhi & Bregu, 2020). The operations of public administration in many countries are now changing due to rapid advancements in computing and digital technologies like Artificial Intelligence (AI). Although there is no universally accepted definition of AI and it is often used interchangeably with deep learning or machine learning, one of the early advocates equated it with intelligent machines (Minsky, 1967). Cerka et al. (2017) considered AI as systems that require minimal human intervention with ability to learn and make decisions. Recently, AI is defined as “a set of technologies that allow machines to learn, reason, interact and deal with uncertainty by themselves” (Ballester, 2021, p.69). The use of AI in public organizations has been at the forefront recently, due to the management of COVID-19 response, but the importance can now be felt in other areas as well. AI-based technologies have been included as one of the critical technologies at various levels of government in many countries (de Sousa et al., 2019) like Brazil to control tax evasion (Faúndez-Ugalde et al., 2020), the United States for the criminal justice (Rizer & Watney, 2018), Singapore to aid contact tracing as the pandemic response (Goggin, 2020) and so on. After 2010, the applications of AI have been incorporated in several public sector functions (Pan, 2016) including armed forces (Ayoub & Payne, 2016), communications (Olsher, 2015), security (Ku & Leroy, 2014), education (Fernandes et al., 2018), transport (Kouziokas, 2017) and public health (Sun & Medaglia, 2019).
AI systems can help governments in multiple ways, in reducing discrimination, human capital costs, time, and freeing up personnel will help governments to help solve complex problems like detecting frauds/abuse, enhancing transparency, fighting human trafficking, preparing for natural disasters, etc. (Newman et al., 2021; Merhi & Bregu, 2020; Wiseman, 2019; Perricos & Kapur, 2019). However, some researchers argue also about the negative impact it may have on public systems like reduction in democratic governance (Young et al., 2019), increase in inequality (Eubanks, 2018), ethical problems (Liaw et al., 2020), and alienating citizens from the government (Byrkjeflot et al., 2018). There are innumerable opportunities for improving the decision-making process if the data is used and analyzed correctly (Iqbal et al., 2020). In spite of heavy investments by governments, Mckinsey (2017) claimed that though $57 billion fraud was discovered by government agencies, $91 billion got completely undetected (Cunningham et al., 2018a). Another report claimed that governments at the global level could save $1 trillion by the use of AI if used correctly (Cunningham et al., 2018b). Deloitte estimated that the use of AI can help governments save 1.2 billion hours of effort and save $41 billion overall (Eggers et al., 2017).