Law as ‘Protection’: Why Algorithmic Decision-Making Matters to Lawyers and the Law
Speaker: Adam Harkens, University of Birmingham, Birmingham Law School, UK
Time: Thursday, 28 October 2021, 10:00 AM CET
Decision-making practices – in both the public and private sector – are increasingly being informed, altered, and automated by algorithmic decision-making (ADM) tools constructed using machine learning techniques. Applications range from predictive advertisement, automated fraud detection, facial recognition, and individual risk assessment for the purposes of informing criminal justice decision making, among more.
Journalism, academic scholarship, and a range of high-profile legal cases have drawn significant attention to the range of dangers and harms associated with the use of these tools: including the potential for biased decision-making and the problems this generates for fairness, a lack of transparency in decision-making, and concerns regarding the potential for automation bias. Accordingly, a number of legislative responses have sought to tackle the challenges posed by these dangers. This includes most notably, the European Commission’s recently proposed Artificial Intelligence Act (AIA), which seeks to introduce legally enforceable standards for both the development and deployment of algorithmic tools within the European Union – wherein ‘AI systems’ are identified and organised by categories of ‘risk’.
Using the law in such a way brings its purpose into stark focus: why is it important that there is a legal response to these dangers? How should we be using the law to respond? And how should we determine what constitutes a ‘successful’ legal response? This talk will explore the crucial role which law (both its rules and principles) plays in protecting individuals against the potential harms of algorithmic decision-making – whose perspective is often at risk of being lost in larger governance debates. For example, despite the potential for ADM tools to generate harm(s) to the fundamental rights of individuals being front and centre of the rationale for the introduction of the AIA, the proposal itself has failed to introduce rights or remedies by which to seek redress for such occurrences. This presentation will discuss a number of ADM tools currently or recently used in practice, in addition to a number of legal responses to their use. In doing so, it will make the case for the need for interdisciplinary understandings of ‘law as protection’ in order to safeguard against the significant number of potential harms that can arise from the use of ADM tools.
The seminar will be broadcasted via Zoom application. Participation is subject to prior registration. The Zoom link of the event will be sent to registered participants via email.
Dr Adam Harkens is a Research Associate at the University of Birmingham, Birmingham Law School (UK) specialising in algorithmic decision-making in the criminal justice system. He is currently working with Professor Karen Yeung (together with 4 German research teams from computer science, neuropsychology, law and political science) on the Volkswagen Stiftung-funded FATAL4JUSTICE? project. This project critically analyses algorithmic decision-making in the criminal justice system from multiple intersecting disciplinary perspectives. Adam and Professor Yeung, as one research team, are responsible for investigating whether the use of ADM systems in the UK and USA is consistent with fundamental principles of the criminal justice system.
Adam’s research interests more generally relate to the socio-legal and theoretical implications of new and emerging technologies on decision-making, data protection law, surveillance, and platform technologies – including legal, technical and ethical modes of regulation and control.
Adam obtained his PhD from Queen’s University Belfast in 2019. His thesis was entitled ‘Law on the Horizon of Algorithmic Governmentality: Legal Innovation, Machine Learning and (Semi-)automated Justice’.
Adam is a regular speaker at academic conferences (e.g. SLSA, CLC, TRILCon, BILETA) and stakeholder roundtables (e.g. RUSI, Bonavero Institute, Royal Irish Academy). He has also recently worked closely with the Société de Législation Comparée and the Dutch Council of State (alongside Dr. Joe Tomlinson, King’s College London) on a project investigating the automation of administrative decision-making across jurisdictions. Adam presented the findings of this project at the Dutch Council of State in April 2019. Adam was a founding member of LawPod, at Queen’s University Belfast, which is a collaborative project bringing together staff and students to explore legal issues in a creative, engaging and accessible way.
The program is supported by the Ministry of Innovation and Technology NRDI Office within the framework of Young Researcher Excellence Program (138965) and the Artificial Intelligence National Laboratory Program.