Endowed Professor 'data and Ai for The Rule of Law'

**Tilburg University | Tilburg Law School and**
**School of Humanities and Digital Sciences, Tilburg University**
is searching for a

**Endowed Professor ‘Data and AI for the Rule of Law’**
**(M/F/x, 0.2 FTE, 5 years)**

Het Wetenschappelijk Onderzoek
- en Datacentrum (WODC), the Ministry of Justice and Security's knowledge center, and Tilburg University are looking to fill a part-time, endowed professor position for 0.2 FTE in the field of data

**Description Chair**:

- Scientific and societal background_
Data
- and AI-intensive systems have great potential to help address major contemporary societal challenges, but they also raise normative challenges themselves. Unless properly designed, trained, implemented, interpreted, and governed, AI systems potentially exacerbate unfair outcomes. At the same time, the use of automated solutions often helps to reveal substantial systemic biases in human decision-making as reflected in data provided for the training of AI systems. It is important that AI-bashing, often accompanied by a lack of knowledge regarding the actual workings of the algorithmic solutions, does not lead to missing out on innovation. Rather, the focus should be on responsible innovation, by developing a socio-technical system where AI supports and improves human-decision making and responsibilities are properly allocated. This is especially important in societal domains where the consequences for people of improper data use can be very serious, such as the justice and security sector.

A vast body of literature has emerged in recent years on AI governance and algorithmic justice, and lawmakers are starting to regulate AI. Yet there is much work still to be done regarding problem definition in the conceptualisation and design stage (which problems can and should AI be used to address?), and in relation to unequal impacts and democratic accountability in the implementation stage (how can AI be used in line with the rule of law and precepts of justice?). In this context, the challenges of data
- and AI-intensive systems call for actionable insights into the legislative and policy initiatives starting to appear in the Netherlands, Europe, and elsewhere.
- and AI-intensive systems are having in practice and how challenges such as bias, fairness, transparency, and accountability play out in specific decision-making processes. Meaningful human oversight and measures to address the uncertainty inherent in all algorithmic systems require both capacity-building and governance frameworks that are currently only in the early stages of development. Moreover, it calls for further guidance both in terms of conceptualisation and implementation.

**Activities**:

- Research questions and goals_
- RQ1: Which actionable insights can be derived from the existing legislative and policy initiatives with respect to the development and use of AI solutions in the domain of justice and security?
- RQ2: Which practical insights can be derived from the study of concrete practices involving data
- and AI-intensive systems in the justice and security sector for AI governance and algorithmic justice?
- RQ3: To what extent can already existing technological, governance, and/or policy solutions be used to address potential points of concern and contribute to the development of data
- and AI-intensive systems for justice and security?
- RQ4: Which algorithmic, governance, and/or policy solutions need to be developed to address the automation needs in the justice and security domain in a way that satisfies the principles of fairness, transparency, and accountability?

In terms of the different academic pillars of activity (research, education, impact, leadership and team spirit), the intended goals to be achieved in fulfilling the mandate of the endowed chair will lie primarily in the pillars of research and impact, with a minor role in educational activities (guest lectures and limited thesis supervision). While the research questions can be approached from a predominant legal or technical perspective, the goal should always be to bridge and bring together the two domains. This can, for instance, result in a combination of theoretical and practice-oriented research (e.g. digital/IT artifact prototyping, use cases, and case-study research), and second, results should be actionable. Significant engagement with AI and data science researchers and practitioners, particularly those working in the domain of algorithmic fairness, AI explainability and human-AI teaming, next to practitioners and experts working in the justice and security sector (including but not limited to law enforcement, immigration bodies, policy-makers, and legislative practitioners) is therefore a must. Results will be published in academic and professional publication venues, policy briefs, or practice guidelines, and where possible made suitable for (guest) lectures in the connected educational programs. We envision to start with one or two PhD projects.
-