A Framework for Understanding Sources of Harm throughout the Machine Learning Life Cycle
This addresses the need for systematic understanding of harm sources in ML to prevent negative societal impacts, though it is incremental as it builds on existing awareness.
The paper tackles the problem of anticipating and mitigating unwanted consequences in machine learning by providing a framework that identifies seven distinct potential sources of downstream harm across the ML life cycle, aiming to improve communication and mitigation strategies.
As machine learning (ML) increasingly affects people and society, awareness of its potential unwanted consequences has also grown. To anticipate, prevent, and mitigate undesirable downstream consequences, it is critical that we understand when and how harm might be introduced throughout the ML life cycle. In this paper, we provide a framework that identifies seven distinct potential sources of downstream harm in machine learning, spanning data collection, development, and deployment. In doing so, we aim to facilitate more productive and precise communication around these issues, as well as more direct, application-grounded ways to mitigate them.