In June, the National Institute of Standards and Technology (NIST) released a draft of a document outlining an approach for identifying and managing bias in artificial intelligence. The team driving responsible AI research at NIST has requested feedback on the document "from people whom AI affects, both from those who create AI systems and also those who are not directly involved in its creation." At Arthur, we are excited to see NIST embrace and elevate the importance of these issues. In our response, we express our support for NIST's efforts as well as address potential gaps in their framework. Read it here.
Figure 1 from NIST’s proposal. Note the arrow from “Deployment” back to “Pre-Design.”