Accountability in Algorithmic Systems: From Principles to Practice
Open Access
- 19 April 2023
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1-4
Abstract
Growing concerns over the societal implications of artificial intelligence has motivated an interdisciplinary push towards mechanisms and tools that hold algorithmic systems accountable. Although there have been considerable strides around defining what it means to hold AI systems accountable, operationalizing those principles have created a barrage of challenges. Researchers, practitioners, and regulators have all raised concerns about the completeness of accountability methods and observed spikes in anxiousness about the potential risk of these tools being manipulated as rubber stamps of approval while harms continue to slip through the cracks. This interactive panel gathers researchers and practitioners with expertise in HCI, Responsible AI, Machine Learning, and Public Policy to critically discuss issues regarding accountability in algorithmic systems to reflect on potential opportunities for re-imagining scalable directions for accountability within these systems.Keywords
This publication has 8 references indexed in Scilit:
- Outsider Oversight: Designing a Third Party Audit Ecosystem for AI Governance2022
- Who Audits the Auditors? Recommendations from a field scan of the algorithmic auditing ecosystem2022
- “I don’t Think These Devices are Very Culturally Sensitive.”—Impact of Automated Speech Recognition Errors on African AmericansFrontiers in Artificial Intelligence, 2021
- Datasheets for datasetsCommunications of the ACM, 2021
- Beyond the individual: governing AI’s societal harmInternet Policy Review, 2021
- When Are Search Completion Suggestions Problematic?Proceedings of the ACM on Human-Computer Interaction, 2020
- Co-Designing Checklists to Understand Organizational Challenges and Opportunities around Fairness in AI2020
- What's Next for AI Ethics, Policy, and Governance? A Global Overview2020