Bias audit
A structured test of a hiring tool that reports, for each protected group, the selection rate and impact ratio, applies the four-fifths rule, names an independent auditor, and states the population and date. A fairness statement that lacks these elements is not a bias audit.
Defined in 2 GAGE programs, which carry 3 distinct definitions of it. The wording above is taught in AI Governance: Applied Mastery.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
An independent review of an automated decision tool's outcomes across different groups, required in New York City under Local Law 144 before an employer may use an automated employment decision tool; a company's own audit history can serve as a checkable fact in a translated talking point.
A structured test of a hiring tool that reports, for each protected group, the selection rate and impact ratio, applies the four-fifths rule, names an independent auditor, and states the population and date. A fairness statement that lacks these elements is not a bias audit.
An independent evaluation of an AI system's outcomes across different demographic or protected-class groups, used to determine whether the system produces disparate impacts. Bias audits are the primary compliance mechanism behind deployer-side fairness laws such as NYC Local Law 144.
Where it is taught
The exact lessons this term appears in. The first 7 topics of every program are free with a free account.
- The hiring-AI problem: bias audits, notices, and the law already watching hiring AI · The EU AI Act: The Executive Map, AI Governance: Applied Mastery
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.