Independent AEDT Bias Audit - NYC Local Law 144 (Disparate Impact / 4-Fifths Analysis)
This is time-sensitive: we need the completed audit by Saturday, June 21 (ET). The scope is intentionally small - we provide the full pseudonymized dataset and the methodology, so the work is primarily the short analysis and a short, signed report (not data wrangling). Please only apply if you can commit to the June 21 (ET) deadline. How our platform works (so you can scope the methodology): We run an online job-fair platform. An AI tool scores each candidate against each job (0–100) based on skills, experience, education, language, and location vs. the job requirements. About the law (context): NYC Local Law 144 requires that AI tools used to screen job candidates undergo an independent bias audit before use. The audit measures impact ratios (selection/scoring rates) across sex, race/ethnicity, and intersectional groups using the 4/5ths rule, and a summary is published. If you've done EEOC adverse-impact analysis, this is the same statistical work applied to NYC's specific reporting format - prior LL144 experience is a plus but not required if your adverse-impact stats are strong. https://www.nyc.gov/site/dca/about/automated-employment-decision-tools.page https://www.osc.ny.gov/state-agencies/audits/2025/12/02/enforcement-local-law-144-automated-employment-decision-tools What we provide: A pseudonymized dataset (no names/PII): candidate ref, sex, race/ethnicity (EEOC categories), match score - a few thousand rows. Optionally, an open-source bias-audit engine you may use, or apply your own methodology. What we need: Impact ratios by sex, race/ethnicity, and intersectional categories (4/5ths rule), aligned with DCWP/LL144 expectations. Appropriate handling of small-sample subgroups; documented methodology, data source, and any excluded records. A signed audit report naming you as the independent auditor of record, plus a publication-ready summary. Ideal background: statistics / I-O psychology / EEO adverse-impact analysis / HR analytics. NYC LL144 familiarity a strong plus. Example of the report that we are looking for https://github.com/aclu-national/tracking-ll144-bias-audits Dover: https://cdn.dover.io/compliance/Dover%20-%20Audit_Result.pdf Bloomberg: https://assets.bbhub.io/company/sites/51/2023/07/20230703-BLP-Bias-Audit-for-AEDT.pdf SmartAssistant: https://perma.cc/YZ8T-CK5M Zoominfo: https://perma.cc/YHV4-XVFC