The use of AI in the public sector often goes hand in hand with ethical concerns, as these new technologies are still new and prone to bias.
The values that are embedded in the use of these technologies and the importance of resilient policymaking to overcome these challenges was discussed during the sixth webinar of the CIVICA PhD Seminar Series on Public Sector Digital Transformation that took place on December 14th with Helen Margetts.
New technologies in public organisations have been largely used to increase efficiency through the automation of tasks previously done by people. Values of efficiency and resource frugality were the backbone of the New Public Management (NPM) reforms of the 1980s, a mentality which has carried over with the recent introduction of complex technologies such as artificial intelligence (AI) in the public sector. Especially in the context of the current health crisis, these tools could prove to be highly useful. For instance, AI could potentially allow for more sophisticated contact tracing methods, personalized treatments or more accurate predictions of where future outbreaks may occur. However, before society can fully take advantage of what these technologies have to offer, it is crucial to foster the trust of citizens through the use of ethical guidelines and good governance.
In order to build citizens’ trust, it becomes increasingly important to better understand the values underlying these technologies and to foster values that help people to comfortably assimilate to using them. As was discussed during the seminar, three sets of values are inherently embedded in public administration: resilience and robustness; honesty, fairness and transparency; and economy and efficiency. However, it is rare that all three can be prioritised equally, and as governments rapidly adapt to the use of new technology and big data, the ethical implications of these new processes often fall to the wayside in favour of efficiency. Especially in times of crisis, it can be easier to focus on efficiency, outcomes and end-goals than to take the time to consider ethics and values. However, ethical issues and their consequences can actually sabotage efficiency in the long run, as they lead to citizen dissatisfaction and distrust in government and new technology. It stands to reason that if people do not trust in these technologies, they will ultimately not use them, and the system will be just as – if not more – inefficient than before.
The challenge of implementing these new technologies in the public sector while also taking into account their ethical implications underscores the importance of resilience in policymaking. Resilience in public service essentially involves having the ability to identify unanticipated events that can disrupt the proper functioning of society and to be prepared to address them when they occur. Focusing on resilience means moving away from parsimonious and efficiency-oriented approaches. One way to do this, as suggested by keynote speaker Helen Margetts, is for data science to move away from doing things humans already know how to do and start doing things humans are not really good at. Moreover, it is important to remember that data is ultimately a reflection of reality and the systems in place. The failures of these technologies can serve to reveal useful information about the systemic biases that exist within society, which, if put to good use, can actually be used to address these issues rather than exacerbate them.
Amidst the current pandemic, the use of AI by governments can potentially offer far more than just efficiency in public service delivery - it might even help save lives. However, as the use and complexity of AI technologies increases, it is crucial for governments to pay more attention to ensuring resilient public policy. Going forward, governments must recognise that citizens’ trust in new technologies is crucial for their successful implementation, and this trust can only be won by prioritising transparency and ethical practices.
This blog post is part of the CIVICA PhD Seminar Series on Public Sector Digital Transformation organised by Hertie School’s Centre for Digital Governance and Bocconi University’s Department of Social and Political Science. The insights highlighted have been based on the sixth seminar session discussion of Helen Margetts’ keynote “Data Intensive Digital Governance in Times of Crisis” as well as her paper “Rethink government with AI”. We would like to thank the participants for sharing their views and ideas.
More literature on this topic:
- Leslie, D. (2020). Tackling COVID-19 through Responsible AI Innovation: Five Steps in the Right Direction. Harvard Data Science Review.
- Lazer, D. M., Pentland, A., Watts, D. J., Aral, S., Athey, S., Contractor, N., & Nelson, A. (2020). Computational social science: Obstacles and opportunities. Science, 369(6507), 1060-1062.
- Bright, J., Ganesh, B., Vogl, T., & Cathrine, S. (2019). Data Science for Local Government. Oxford Internet Institute.