The potential and actual use of Big Data (BD) applications affects aspects of the theoretical and practical considerations of decision-making, learning and process optimisation in the public sector both theoretically and practically (Giest 2017). The impact of BD is driven not only by the data revolution but also the accompanying development of new technologies (e.g. AI-driven technologies, blockchain, distributed ledger technology…etc.) and advanced analytics (e.g. machine learning algorithms). Many public administration organisations around the world have already started to deploy AI-powered interfaces for citizen response systems, legal adjudication processes, fraud detection, and infrastructure planning.  Lember et al. (2019) argue that BD creates the opportunity to go from descriptive analysis to predictive and even prescriptive analysis and consequential policy development. Several other authors stated that better use of BD can result in benefits to the public sector (Maciejewski, 2017; Mergel, et al. 2016). Nevertheless, when it comes to describing the actual applications and advantages, authors diverge considerably in their approach and conclusions (Pencheva, et al. 2018). As Janssen and Kuk (2016) underline, the design and training of the algorithms that exploit BD are not neutral, not free from human interferences and not free from biases. Ensuring transparency and accountability is according to the authors a critical success factor. Furthermore, Klievink et al (2017) found in the case of the Netherlands, the public sector organisations may be technically capable of using BD, but they will not significantly gain from BD if the applications do not fit their organisations and statutory tasks.

DIGI4FED starts from those challenges and aims to develop a governance design that serves the internal administrative and public service processes of the Belgian federal government; a governance design that is embedded in the open governance ecosystem and makes full use of the potential offered by big data (BD) and its application via artificial intelligence (AI) and blockchain technology (BCT).

In particular, three contextual factors define the context by which DIGI4FED is influenced. The first factor is the growing attention for the potential impact of BD and AI on traditional government information processes. Indeed, in recent years the concepts of BD and AI led to rethinking the design of traditional government information processes in the Belgian public administrations (Heijlen et al. 2018). There are, however, several challenges related to the exploitation of BD through AI, such as the risk for goal displacement due to a potential overreliance on performance metrics and predictive analytics as well as a potential lack of precision in those performance metrics. Indeed, the objective function to be optimised through AI heavily depends on the data used to train the algorithms. Furthermore, individuals may lose trust in public administration due to a potential lack of transparency and accountability in how policy is made and the high level of required technical expertise (Lavertu, 2015).

The second factor is the growing expectation of society from public administrations, to adopt new technological means to advance efficient and effective governance and public service delivery whilst ensuring the core democratic and moral values are not lost out of sight (EC, 2013). Here the developments in two IT technologies, AI and BCT, are particularly important for the effective uptake of the BD solutions by the public administration. On the one hand, AI enables public administrations to gain new policy insight and to develop improved policies and action through effective exploitation of the public data. On the other hand, BCT can support the pursuance for public values such as transparency and democratic accountability (Ølnes, et al., 2017) by giving control of the data to users. As such BCT has the potential to enhance the usage of private data in public processes without undermining the ethical and legal rights on user data.

A final and third factor concerns the Belgian federal administration itself. Although in the past, several steps were taken towards the digital transformation of the Belgian federal state, challenges remain. The BELSPO BRAIN-be projects “FLEXPUB” and “PSI-CO” have demonstrated that the federal level has certain requirements and faces challenges concerning digitalisation and innovation. A further digitalisation is welcomed, both in and outside the administration.  Identified challenges include trust and equity embedded in digital tools, transparency of data and data use, the human dimension of digital judgments, technical difficulties, human and cultural factors, a gap between ambition and reality, and an imbalance between in- and outsourcing.    

Taking into account the abovementioned three contextual factors and the scope of the call, the main objective of DIGI4FED is to understand ‘how (big) data can be used in the Belgian federal administration system to enable better public provision through new technologies such as AI and blockchain technology?

Share this page