Document Actions
Graduate Center Colloquium
Beyond Opening the Black Box: Operationalizing Audits and Assessments for Consequential AI
Categories |
|
---|---|
When |
May 12, 2021 from 01:00 pm to 02:00 pm |
Hosting organization | Graduate Center Computer Science Department |
Speaker | Joshua Kroll |
Speaker Information | Joshua A. Kroll is an Assistant Professor of Computer Science at the Naval Postgraduate School studying the relationship between governance, public policy, and computing systems. His University of Pennsylvania Law Review paper "Accountable Algorithms" received the Future of Privacy Forum's Privacy Papers for Policymakers Award in 2017. Joshua helped to create and remains involved with the ACM Conference on Fairness, Accountability, and Transparency. His previous work spans accountable algorithms, cryptography, software security, formal methods, Bitcoin, and cybersecurity policy. Joshua holds a PhD in computer science from Princeton University and a bachelor's degree in Physics and Mathematics from Harvard College. |
Where | On-line |
Contact Name | Prof. Raja |
Contact Email | anita.raja@hunter.cuny.edu |
Add event to calendar |
vCal iCal |
Automation is not an unalloyed good: computational systems cause real harms in practice, from biased outcomes to lethal accidents. Principles for responsible and ethical AI abound, but risk becoming empty promises without a bridge from abstract values to technology. How will accountability be realized? We explore how system-level assessments support oversight regimes for consequential AI. What standards of oversight readiness reflect adequacy? To what extent do control structures from other domains inform assessment requirements in AI systems? We systematize these challenges, the paths to managing them, and open research problems in realizing the goal of trustworthy and responsible AI.