Conference Agenda

Generative AI has surged recently, promising to transform production processes in the public sector too. This technology can assist or manage various information processing tasks within government organizations. A UK study shows over 20% of public service professionals use Gen AI (Bright et al., 2024). However, critics highlight rising sustainability costs, like energy and water usage. These costs are often overlooked (Crawford, 2024). Governments leveraging GenAI must find ways to mitigate these negative impacts, but guidance on managing sustainability costs is limited. This paper proposes a method for government organizations to handle these sustainability costs effectively.

Managing GenAI is complex due to a threefold uncertainty. Regulation uncertainty refers to the permissions to use GenAI in government, technological uncertainty concerns the development of the technology, and behavioral uncertainty encompasses how users will respond to the introduction of GenAI. This paper, therefore, introduces a ready-to-use decision support model for the project, organization and national levels. This decision support model builds upon methods for scenario planning, top-down modeling, and design thinking. The result is a method that project managers, CIOs of organizations and policy-makers can use to support their GenAI strategy. The method allows for mapping their expected growth in GenAI use, modelling what this will mean for concomitant sustainability costs, and choosing from a ‘menu’ of mitigation methods to reduce the sustainability costs appropriately. The paper uses interviews with frontrunners in GenAI within government to outline scenarios. These scenarios serve as the basis for top-down modeling, which estimates energy use, water use, and CO2 emissions. The information from the interviews and the top-down modeling is then utilized in a design workshop where mitigation strategies are developed.

We find that GenAI is being developed internally by governments, and the potential of innovative GenAI solutions is large. Externally developed GenAI is, moreover, used despite this not being allowed. This means that external GenAI use falls under ‘Shadow IT’’. GenAI can develop in three ways in the coming years. Development can, for instance, be slow and incremental, or a radical transformation can take place. What we expect, however, is a transition toward more integration of GenAI in some, but not all, government processes. Concerning sustainability, we foresee that the power consumption of GenAI within the government could reach about 14.6, 27.2, or even 53.3 GWh on an annual basis in three years, respectively. By comparison, the total electricity consumption of government data centers was still at 128 GWh per year in 2020. Derived from these costs, CO2 emissions and water consumption were calculated. The projected power costs of GenAI within the government translate to 3.9, 7.4 and 14.4 million kilograms of CO2 emissions and 13.7, 25.4 and 49.8 million liters of water consumption, respectively. Mitigation strategies are possible at the organizational and national level. At the organizational level, attention to sustainability in digitalization policies is crucial, as well as monitoring and making deliberate choices for sustainable alternatives. On the national level, central steering and increasing awareness about the costs of sustainability are essential.

Environmental and climate-related issues are becoming increasingly severe and the need to develop more sustainable societies has risen significantly. Although public awareness of environmental impacts has improved, this has not sufficiently translated into sustainable patterns of behaviour as many people still ignore environmental consequences even when aware of their actions (Bolderdijk et al., 2018). As such, pro-environmental behavioural change in citizens is a pressing challenge for governments compelling them to prioritize sustainable development in policy making and implementation (Lucas et al., 2008).

To better address the gap between public awareness and the actual adoption of sustainable behaviours, digital nudging can encourage more eco-friendly choices. Nudging is a cost-effective and scalable approach to encouraging pro-environmental behaviours and has gained widespread attention (Byerly et al., 2018). Nudges can shape decision-making environments subtly, making sustainable options more accessible and appealing thereby promoting long-term behavioural change while respecting autonomy (Thaler and Sunstein, 2008). Moreover, the influence and reach of the public sector provide a unique opportunity for large-scale nudging implementation.

This paper presents the results from a Structured Literature Review (SLR) that evaluates the ‘state of the art’ regarding the potential of nudging to affect pro-environmental behaviour change in an eGovernment context.

Digital nudging has been effectively enhancing user engagement in the private sector through strategic design. Simplifying information presentation (Johnson et al., 2012), like Apple's structured pricing or Amazon’s categorized listings, reduces decision-making complexity while personalized data-driven recommendations like Netflix's tailored content promotes satisfaction and participation (Gomez-Uribe and Hunt, 2015).

Whilst wildly applied in the private sector, the exploration of digital nudging in the public sector remains an emerging field. Additionally, existing literature presents mixed results about the long-term effects of digital nudges on changing environmental behaviour (Zimmermann et al., 2021). Implementing digital nudging in e-Government faces several challenges, including ethical concerns about autonomy, transparency, and trust, particularly the risk of perceived manipulation (Michaelsen, 2024, Oliver, 2013, Rebonato, 2012, Hausman and Welch, 2010). Ethical questions revolve around whether the “digital nudges” respect autonomy or subtly pressure choices, thereby undermining their trust in government. Privacy is another critical issue due to the extensive data use required for personalization (Hettler et al., 2024). Legal uncertainties and legitimacy concerns hinder digital nudging practices, as existing legal frameworks may not clearly define or support its use (Alemanno and Spina, 2014).

Additionally, challenges in design, implementation, resource constraints, and scaling across diverse contexts can reduce the potential impact (Ranchordás, 2020). Government overreach is also a concern as the more a government designs subtle and psychologically influential choice structures that are difficult to bypass, the greater the concern for potential threats to individual autonomy (Bovens, 2009).

This systematic literature review follows the Okoli and Schabram (2015) guidelines for Information Systems research (Okoli and Schabram, 2015) and addresses a critical gap by providing a comprehensive synthesis of existing knowledge on digital nudging within e-Government for promoting pro-environmental behaviours. The article also provides a framework to guide essential areas of future research and guidelines for the implementation of this important technology.

As digital technologies reshape the global economy, governments are increasingly adopting digital platforms to improve service delivery, transparency, public satisfaction, and to enhance engagement with both citizens and businesses. The existing studies have generally shown a positive and significant correlation between digitalization (of government) and economic growth in both developed and developing countries. However, these studies often adopt a limited perspective, overlooking a more comprehensive and holistic approach. On the one hand, they overlook the fact that the rapid surge in digitalization is transforming the complex structure of the digital society, which comprises interactions between governments, businesses, and citizens. On the other hand, they fail to recognize that in modern realities, pure economic growth should be redefined as economic sustainability, which, besides economic development, includes social inclusiveness and environmental responsibility. Namely, pursuing social and particularly environmental issues has often been identified as a key obstacle to economic growth, which remains a significant challenge. Digitalization has the potential to address this issue by enhancing governance quality through the streamlining of administrative processes, improved access to real-time data, and more transparent and participatory decision-making, thereby enabling governments to design and implement policies that more effectively balance economic, social, and environmental priorities.

Therefore, this paper investigates the role of e-governance in driving economic sustainability within the context of a digital society in the European Union (EU). The methodology is structured around a regression analysis to explore the relationship between e-governance and key indicators of economic sustainability, therefore a multiple linear regression model will be used to analyze the relationship between e-governance adoption and economic sustainability. Data will be collected from a variety of sources, including government reports, EU reports, OECD reports, global e-governance indicators, and economic sustainability indicators such as GDP growth, income inequality, and carbon emissions. The data will cover 27 EU countries over 10 years (2014-2024) with varying levels of e-governance adoption, spanning both developed and developing nations. The data will be analyzed using statistical software, such as SPSS or R, to estimate the regression coefficients and evaluate the significance of the relationships.

The proposed research question will contribute to new empirical evidence on the impact of digitalization on economic sustainability. E-governance plays a significant role in promoting economic sustainability by improving efficiency and cutting costs for both governments and businesses, driving economic growth through better public service delivery and innovation. Socially, e-governance increases access to services, especially in remote areas, and encourages citizen participation, making governments more accountable. It also promotes social equity by providing equal access to services and enhances public health and education through digital platforms. Environmentally, e-governance helps reduce paper usage and waste, supporting sustainability, and lowers carbon emissions by decreasing the need for physical transportation and infrastructure. Overall, e-governance contributes to economic growth, social inclusion, and environmental preservation. The expected results would emphasize that e-governance, when implemented effectively, acts as a catalyst for sustainable economic development, driving transparency and efficiency, while also addressing societal and environmental challenges.

As governmental organisations digitalise their operations, the environmental impact of ICT is becoming a rising concern. While e-Government enhances efficiency and service delivery, it also adds to increased energy consumption and greenhouse gas emissions. This paper investigates the need for a systematic approach to Green e-Government by proposing a five-pillar model for improving digital governance sustainability. This study sheds light on how green practices are implemented by conducting a qualitative study of five Dutch public organisations. The findings show that, while there are many Green ICT initiatives, they do not integrate into a larger green e-Government strategy. In addition from its results from empirical research, this study provides a theoretical contribution by bringing Green e-Government to a more mature stage. The five-pillar model is a structured framework that incorporates sustainability into digital governance, laying the foundation for future study and policy development. By bridging the gap between Green ICT and e-Government, this work adds to theoretical and practical conversations about Green E-Government. It provides an actionable framework for policymakers and public administrators to ensure that digital transformation is consistent with environmental sustainability objectives.