This document focuses on employment practices that may have a disparate impact on the basis of race, color, religion, sex, or national origin, and potentially in violation of Title VII. The information it offers is nonbinding and is intended only to provide clarity regarding existing requirements under the law.
The document provides definitions of key terms regarding automated systems and AI including:
Algorithm: a set of instructions that a computer can follow to accomplish some end, such as decision-making at various stages of employment (hiring, performance evaluation, promotion, and termination).
Artificial intelligence (AI): defined by Congress as “a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.”
The document also provides a Q&A on the following:
Could an employer’s use of an algorithmic decision-making tool be a “selection procedure”?
Can employers assess their use of an algorithmic decisions-making tools for adverse impact in the same way they assess more traditional selection procedures?
Is an employer responsible under Title VII for its use of algorithmic decision-making tools even if they were designed or administered by another entity?
Does compliance with the four-fifths rule (the ratio used to compare selection rates between job candidate groups) guarantee that a particular employment procedure does not have an adverse impact?
If an employer discovers use of an algorithmic decision-making tool would have an adverse impact, may it adjust the tool, or decide to use a different tool, to reduce or eliminate that impact?