Outsider Oversight: Designing a Third Party Audit Ecosystem for AI Governance

By Inioluwa Deborah Raji, Peggy Xu, Colleen Honigsberg, Daniel E. Ho. 2022

180 citations recorded by the source corpus.

Much attention has focused on algorithmic audits and impact assessments to hold developers and users of algorithmic systems accountable. But existing algorithmic accountability policy approaches have neglected the lessons from non-algorithmic domains: notably, the importance of third parties. Our paper synthesizes lessons from other fields on how to craft effective systems of external oversight for algorithmic deployments. First, we discuss the challenges of third party oversight in the current AI landscape. Second, we survey audit systems across domains - e.g., financial, environmental, and health regulation - and show that the institutional design of such audits are far from monolithic. Finally, we survey the evidence base around these design components and spell out the implications for algorithmic auditing. We conclude that the turn toward audits alone is unlikely to achieve actual algorithmic accountability, and sustained focus on institutional design will be required for meaningful third party involvement.

Subfields: Information Systems, Safety Research, Strategy and Management, AI Ethics, Fairness & Society

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems

1. Jonas Schuett
2. Irene Solaiman
3. Andrew Smart
4. Luciano Floridi
5. Jakob Mokander
6. Hannah Rose Kirk

In Safety Research

1. Jonas Schuett
2. Irene Solaiman
3. Andrew Smart
4. Luciano Floridi
5. Jakob Mokander
6. Negar Rostamzadeh
7. Hannah Rose Kirk
8. Jessica Morley

In Strategy and Management

1. Andrew Smart
2. Luciano Floridi
3. Negar Rostamzadeh

In AI Ethics, Fairness & Society

1. Jonas Schuett
2. Irene Solaiman
3. Kathryn Henne
4. Andrew Smart
5. Luciano Floridi
6. P. Nicholas
7. Jakob Mokander
8. Gurleen Virk

Citations in the corpus

• The Gradient of Generative AI Release: Methods and Considerations (2023, background): “Formal audits alone cannot be the only insight or governance of a system [76].”
• Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools (2024, unknown): “If only well-heeled actors can even evaluate the risks of AI systems, claims of functionality (Raji et al., 2022) and that AI can improve access to justice may be quite overstated (Bommasani et al., 2022; Chien et al., 2024; Perlman, 2023; Tan et al., 2023).”
• Sociotechnical Harms of Algorithmic Systems: Scoping a Taxonomy for Harm Reduction (2022, background): “As such, it presents navigational challenges for practitioners seeking to comprehensively evaluate a system for potential harms [178, 179, 182], particularly for large generative models that perform different tasks across many use cases.”
• Sociotechnical Harms: Scoping a Taxonomy for Harm Reduction (2022, background): “Researchers have also developed “ethics methods” [131] for practitioners to identify and mitigate harms, including statistical assessment [116, 126], software toolkits [28], and algorithmic impact assessments and audits [157] that offer significant value for practitioners.”
• Auditing large language models: a three-layered approach (2023, background): “, who should conduct the audit and how to ensure post-audit action [13].”