Racially Biased AI Model: A Case for Responsible AI Governance

October 22, 2021

The publicly accessible Delphi AI, allegedly biased in ethical advice due to training via Reddit threads, highlights the importance of trustworthy AI governance and safe data use in language-based models. This AI incident maps to the Govern function in HISPI Project Cerebellum Trusted AI Model (TAIM). Ready to help shape responsible AI? JOIN US
Alleged deployer
allen-institute-for-ai
Alleged developer
allen-institute-for-ai
Alleged harmed parties
minority-groups

Source

Data from the AI Incident Database (AIID). Cite this incident: https://incidentdatabase.ai/cite/146

Data source

Incident data is from the AI Incident Database (AIID).

When citing the database as a whole, please use:

McGregor, S. (2021) Preventing Repeated Real World AI Failures by Cataloging Incidents: The AI Incident Database. In Proceedings of the Thirty-Third Annual Conference on Innovative Applications of Artificial Intelligence (IAAI-21). Virtual Conference.

Pre-print on arXiv · Database snapshots & citation guide

We use weekly snapshots of the AIID for stable reference. For the official suggested citation of a specific incident, use the “Cite this incident” link on each incident page.