Permanent project revision

Local Law 144: Bias Audits for Hiring Algorithms

Revision 2026-08-20-source-dates · effective 2026-07-26

Current revision. This is the snapshot used by the live entity page.

unreviewed

Created the first structured Evidence Passport with 1 claim and 6 source dossiers.

sha256:387f09555b0bae96688a16fb7835d4b4d5d9f995ad789a3e38d348d3bf258dc1

All revisions → Current entity page →

Change set

add /evidencePassport — Added claim-level evidence mapping and source dossiers.
add /revisionHistory — Established a permanent, hash-addressed baseline revision.

1: of 1 claims mapped
6: source dossiers
4: publication dates recorded
1: archives linked
0: exact locators
0: content captures hashed

Claims, one by one

The text below is copied verbatim from the published record. Each claim names the source dossiers that carry it; caveats stay attached to the claim they qualify.

Claim 1 clm-4f0d15881428c77f quantitative

18 of 391 employers checked had posted a bias audit report and 13 a transparency notice

Published field: impact · verbatim
Mapping: mapped · no interpretation or interpolation added by this passport
What the mapping says: The entry states researchers who checked 391 employers in autumn 2023 found 18 published audit reports.
Caveat: No claim-specific caveat recorded

Sources carrying this claim

Wright et al., Null Compliance: NYC Local Law 144 and the Challenges of Algorithm Accountability (ACM FAccT 2024)

Source dossiers and revision state

This layer records what is known about each source, what is still missing, and the exact snapshot this page is rendering. A missing date, archive or hash stays visible as missing.

unreviewed Revision 2026-08-20-source-dates effective 2026-07-26 · created 2026-08-20

Awaiting operator and domain-expert review; publication of this revision implies neither.

sha256:387f09555b0bae96688a16fb7835d4b4d5d9f995ad789a3e38d348d3bf258dc1

Revision history → Permanent version link → Machine-readable record → Evidence Constitution →

1: of 1 claims mapped
6: source dossiers
4: publication dates recorded
1: archives linked
0: exact locators
0: content captures hashed

Governance facts carried by the record

Operator: New York City Department of Consumer and Worker Protection (DCWP)
Actor types: city-government
Funding model: public
Funding description: Not recorded
Cost: Not recorded
Ownership: Not recorded
Decision rights: Not recorded
Exit conditions: Not recorded

6 source dossiers

NYC Department of Consumer and Worker Protection — Automated Employment Decision Tools nyc.gov · background

Open original source ↗

Stable source ID: src-3786d4e66e97f74d
Publisher: Name not classified · nyc.gov
Type / language: web-page · language not classified
Published: Not recorded
Updated: Not recorded
Source last checked: No source-level check date
Record last verified: 2026-07-26
Dossier created: 2026-08-20
Retrieved copy: No local capture recorded
Archive: No archived copy linked
Content hash: No captured content hash
Locator: Open the recorded source note
The regulator's own page: enforcement begins 5 July 2023; the tool must have had a bias audit within one year of use, information about the audit must be publicly available and a summary of results posted, and notice must be given 10 business days before an AEDT is used.
Descriptive only — not page/paragraph exact
Claims linked: Background source; no baseline claim points here

NYC Rules — Automated Employment Decision Tools (DCWP rulemaking) rules.cityofnewyork.us · background

Open original source ↗

Stable source ID: src-83e8ed42b0464bab
Publisher: Name not classified · rules.cityofnewyork.us
Type / language: web-page · language not classified
Published: Not recorded
Updated: Not recorded
Source last checked: No source-level check date
Record last verified: 2026-07-26
Dossier created: 2026-08-20
Retrieved copy: No local capture recorded
Archive: No archived copy linked
Content hash: No captured content hash
Locator: Open the recorded source note
The implementing rule: defines a covered tool as machine learning, statistical modelling, data analytics or artificial intelligence whose output is used to substantially assist or replace discretionary decision making, and sets out selection rate, scoring rate and impact ratio calculated by sex category, race/ethnicity category and intersectional categories.
Descriptive only — not page/paragraph exact
Claims linked: Background source; no baseline claim points here

Wright et al., Null Compliance: NYC Local Law 144 and the Challenges of Algorithm Accountability (ACM FAccT 2024) arxiv.org · 1 linked claim

Open original source ↗

Stable source ID: src-d173e9be1278798e
Publisher: Name not classified · arxiv.org
Type / language: web-page · language not classified
Published: 2024-06-03
Updated: Not recorded
Source last checked: No source-level check date
Record last verified: 2026-07-26
Dossier created: 2026-08-20
Retrieved copy: No local capture recorded
Archive: Open archived copy ↗ · 2026-03-05
Content hash: No captured content hash
Locator: Open the recorded source note
Peer-reviewed source for the compliance figure: 155 student investigators recorded 391 employers between 24 October and 9 November 2023; 18 posted audit reports and 13 posted transparency notices. Also the source for the 'null compliance' problem, employers holding near-total discretion over whether their system is in scope. Published at FAccT '24, DOI 10.1145/3630106.3658998.
Descriptive only — not page/paragraph exact
Claims linked: clm-4f0d15881428c77f

Groves, Metcalf, Kennedy, Vecchione and Strait, Auditing Work: Exploring the New York City algorithmic bias audit regime (ACM FAccT 2024) arxiv.org · background

Open original source ↗

Stable source ID: src-4ecd11298d6b0087
Publisher: Name not classified · arxiv.org
Type / language: web-page · language not classified
Published: 2024-02-12
Updated: Not recorded
Source last checked: No source-level check date
Record last verified: 2026-07-26
Dossier created: 2026-08-20
Retrieved copy: No local capture recorded
Archive: No archived copy linked
Content hash: No captured content hash
Locator: Open the recorded source note
The methodological counterweight, from interviews with auditors and practitioners: the law fails to define AEDTs and independent auditors clearly, rests on a flawed transparency-driven theory of change, and industry lobbying narrowed the definition of a covered tool. Also the source characterising LL 144 as the first algorithm auditing system for commercial machine-learning systems.
Descriptive only — not page/paragraph exact
Claims linked: Background source; no baseline claim points here

Office of the New York State Comptroller — Enforcement of Local Law 144, Automated Employment Decision Tools (2 December 2025) osc.ny.gov · background

Open original source ↗

Stable source ID: src-4dda8f61b88a4285
Publisher: Name not classified · osc.ny.gov
Type / language: web-page · language not classified
Published: 2025-12-02
Updated: Not recorded
Source last checked: No source-level check date
Record last verified: 2026-07-26
Dossier created: 2026-08-20
Retrieved copy: No local capture recorded
Archive: No archived copy linked
Content hash: No captured content hash
Locator: Open the recorded source note
Independent state audit covering July 2023 to June 2025: DCWP's system for enforcing the law is ineffective, only two AEDT-related complaints were received in the whole period, and the complaint intake process does not reliably route complaints to DCWP.
Descriptive only — not page/paragraph exact
Claims linked: Background source; no baseline claim points here

New York State Comptroller press release on the DCWP audit (December 2025) osc.ny.gov · background

Open original source ↗

Stable source ID: src-dd580252ec17e825
Publisher: Name not classified · osc.ny.gov
Type / language: web-page · language not classified
Published: 2025-12-02
Updated: Not recorded
Source last checked: No source-level check date
Record last verified: 2026-07-26
Dossier created: 2026-08-20
Retrieved copy: No local capture recorded
Archive: No archived copy linked
Content hash: No captured content hash
Locator: Open the recorded source note
Source for the specific enforcement numbers: DCWP reviewed 32 employers' and vendors' websites and identified one instance of likely non-compliance, while state auditors reviewing the same companies found 17; of 12 test calls to 311 only three reached DCWP.
Descriptive only — not page/paragraph exact
Claims linked: Background source; no baseline claim points here

Known limits of this passport

Publisher names and source languages have not yet been classified; domains are recorded without guessing.
Existing source notes are descriptive locators, not page- or paragraph-exact citations.
A missing archive, publication date or content hash is shown as null rather than inferred.
The project-level evidence grade ranks the strongest source in the entry; it is not a truth score for every claim.