Inspiration
I proposed this application because I am interested in how we can ensure the digitalisation of the public sector remains human-centric. I first became interested in this idea after witnessing the life-altering consequences left in the wake of the Dutch childcare benefit scandal while I was living in the Netherlands. The Dutch tax authorities had been relying on the SyRI welfare fraud detecting algorithm for a period of six years, during which the erroneous classification of thousands of individuals brought upon them financial and personal ruin. Tens of thousands of families — often with lower incomes or belonging to ethnic minorities — were pushed into poverty because of exorbitant debts to the tax agency. The algorithm was abandoned after the Dutch Data Protection Authority found that it had illegally used nationality as a data variable, as well as whether someone, or their family relations, might be connected to a non-EU country. Upon witnessing the repercussions of the Dutch welfare scandal, I started thinking about the type of AI tools governments readily adopt for the sake of efficiency without a nuanced understanding of the technical limitations of machine learning. Governments seem to trust that if an AI system works well in the private sector, it will perform just as well in the public sector.
Throughout my academic journey, I spent time exploring the relationship between the digital welfare state and vulnerable citizens. The public sector has widely embraced AI-powered tools as a cost-cutting and efficiency-enhancing solution. However, bureaucratic optimisation has come at a cost. I argue that the nature and meaning of ‘optimisation’ in the public sector are changing to accommodate the values that drive private sector machine learning. The goals of public sector digitalisation will thus shift from human-centric principles and values such as autonomy, dignity, and non-discrimination towards economic metrics of success.
By reimagining the public procurement process as a decentralised app, I hope to address two key concerns. Firstly, private sector metrics of success are imported into the public sector via the AI pipeline. The public sector uses AI for the same reasons as private companies—to increase efficiency and speed of transactions. However, one may argue that public bodies must first and foremost abide by principles of good governance, rather than prioritise economic optimisation. Secondly, reliance on off-the-shelf AI tools discourages public bodies from developing internal technical capabilities.
What it does
I propose the development of a decentralised application designed to streamline the public procurement process. This Daap will govern the integration of privately developed AI tools into the public sector, addressing key implications of current public-private partnerships. The proposed Daap will provide its participants with deeper insight into the AI pipeline and allow public officials more control over what type of innovation is allowed into the public sector. The proposed Daap will enhance efficiency, transparency, and accountability. Secondly, the Daap will encourage competition not solely based on cost and functionality but also the societal impact of an AI system. As a result, public officials will be able to select software tools that are geared toward a sustainable, equitable, and green digital economy.
The proposed daap operates as a decentralised blockchain based application that will help public sector organisations when buying supplies and services or tendering public works. The enhanced transparency and open-access structure of the app will enable public sector organisations to select software tools not solely based on cost and functionality, but also on potential governance and human rights risks. The goal is to address the imbalance of power and capabilities which defines the existing public-private partnerships. Namely, public sector organisations have grown reliant on the private sector to drive innovation, which in turn excludes governments from innovation. In this daap, public organisations can pick applications that will gear the development of AI towards public values, such as a sustainable, equitable and green digital economy.
Built With
- blockchain
- permissioned
- procurement