Can Algorithms Bend the Arc Toward Justice?

Meeting Summary:  This virtual SFI workshop on Algorithmic Justice is organized by the SFI-UNM Interdisciplinary Working Group on Algorithmic Justice and is the fourth in a series sponsored by our NSF-funded program “Foundations of Intelligence in Natural and Artificial Systems.” Law, Computer Science, Philosophy and the Social Sciences have attempted to articulate fundamental questions about the nature of justice in algorithms and artificial intelligence (AI). This workshop will connect top scholars and practitioners in these fields to understand how algorithms in criminal justice, healthcare, employment, and other systems can either exacerbate or mitigate bias, unfairness, and injustice in society. Ultimately, our goal is to advance algorithms that transform notions and implementations of justice to create a more ethical society. An interdisciplinary group of speakers will address questions about the role of algorithms in a just society, including: 

- What are definitions of justice and fairness in CS, social science, law and philosophy? How can we expand the notions of justice beyond utilitarian definitions that are often considered in evaluating algorithmic justice?

- What kind of transparency, oversight, regulation, incentives, or controls could lead to algorithms/AI that mitigate bias and unfairness in existing systems?

- How can AI be embedded in existing systems to better reflect, reveal, correct, or mitigate existing biases and injustices?

- What frameworks or unifying theoretical perspectives enabled by algorithms or AI can transcend current structures and practices in human decision-making to generate more just outcomes in interconnected, complex socio-technical systems?