What’s in a name? A survey of strong regulatory definitions of automated decision-making systems
The Electronic Privacy Information Center (EPIC) emphasizes the necessity of adopting broad regulatory definitions for automated decision-making systems (ADS) to ensure comprehensive oversight and protection against potential harms.
In the article “What’s in a Name? A Survey of Strong Regulatory Definitions of Automated Decision-Making Systems,” EPIC Counsel Ben Winters discusses the lack of consensus on defining ADS and artificial intelligence.
The piece highlights the importance of expansive definitions to effectively regulate these technologies, noting that even simple tools can cause significant harm if not properly managed. The article surveys various definitions proposed by scholars and legislative bodies, including:
- Scholarly Definitions: Rashida Richardson defines an “Automated Decision System” as any computational tool designed to automate, analyze, aid, augment, or replace government decisions, impacting areas such as opportunities, access, rights, and status.
- Legislative Definitions:
– D.C. Stop Discrimination by Algorithms Act: Defines “algorithmic eligibility determination” as decisions based significantly on algorithms utilizing machine learning or similar techniques to determine individual eligibility for important life opportunities.
– Washington State SB 5116: Describes an “automated decision system” as electronic software or processes intended to automate, aid, or replace decision-making impacting residents’ welfare or rights, including tools that generate scores, predictions, classifications, or recommendations.
Share this Resource:
