Service

Independent AEDT bias audit

NYC Local Law 144 requires an annual, independent bias audit of any automated employment decision tool (AEDT) used to screen, rank, or score candidates for a New York City role. We perform that audit, sign it independently, and deliver everything you are required to publish.

What we test

For each tool we calculate the selection rate (or scoring rate for tools that output a score) for every demographic category the law covers: sex, race and ethnicity, and the intersectional combinations of the two. We then compute the impact ratio, the rate for each group divided by the rate of the highest-scoring group.

Any impact ratio below 0.80 trips the four-fifths rule, the long-standing federal threshold for potential adverse impact. Example: if men are selected at 60% and women at 40%, the impact ratio for women is 0.67, which falls below the threshold and is flagged in the report.

What you receive

• 1. A full bias-audit report with selection or scoring rates, impact ratios, and intersectional breakdowns, plus a plain-English interpretation.
• 2. The public summary in the exact format the law requires you to post, including the audit date, the data source, and the count of categories below the 0.80 impact ratio.
• 3. A candidate-notice template for the 10-business-day advance notice the law requires.
• 4. A named, independent sign-off. We have no financial interest in your tool or your company, which is what the law requires.

The data we need

We work from aggregated selection or scoring data by demographic category wherever possible, so we rarely need raw applicant records. Where historical data is unavailable, we run the audit on test data and disclose that in the report. All data is transferred securely, used only for the audit, and retained no longer than necessary.