NYC Local Law 144 - AI Hiring Bias Audit Compliance

NYC Local Law 144, effective July 5, 2023, is one of the first laws in the United States to regulate the use of AI in hiring decisions. The law requires employers and employment agencies in New York City to conduct annual independent bias audits of automated employment decision tools (AEDTs) before using them, publicly post audit results, and provide notice to candidates and employees subject to AEDT evaluation.

The law was enacted in response to growing concerns about algorithmic discrimination in hiring, where AI tools trained on historical data may perpetuate or amplify existing biases related to race, ethnicity, and gender. By mandating transparency and independent auditing, Local Law 144 aims to ensure that AI-driven hiring tools do not unfairly disadvantage protected groups.

While Local Law 144 applies specifically to New York City, it has established a template that other jurisdictions are following. The Colorado AI Act, Illinois AI Video Interview Act, and proposed federal legislation all draw on principles first codified in Local Law 144.

Organizations using AI in hiring and talent management can leverage Areebi to implement governance controls that satisfy Local Law 144 and prepare for similar requirements in other jurisdictions. The platform's policy engine, audit trails, and compliance dashboards provide the infrastructure for ongoing AEDT governance.

Local Law 144 defines an AEDT as any computational process derived from machine learning, statistical modeling, data analytics, or artificial intelligence that issues a simplified output - including a score, classification, or recommendation - used to substantially assist or replace discretionary decision-making for employment decisions.

Key aspects of the definition:

• "Substantially assist" means the tool's output is weighted more heavily than other factors in the decision, or is used to overrule conclusions derived from other factors (including human decision-making)
• Employment decisions include hiring and promotion decisions within New York City
• Tools that do not generate a simplified output (e.g., tools that generate only unstructured text) may not qualify as AEDTs under the law

Common examples of AEDTs include AI-powered resume screening tools, candidate ranking systems, video interview analysis platforms, skills assessment scoring tools, and AI-driven promotion recommendation engines.

The core requirement of Local Law 144 is the annual independent bias audit. Key requirements include:

• Independence: The audit must be conducted by an independent auditor who is not involved in using or developing the AEDT and whose compensation is not based on audit results
• Annual frequency: The bias audit must be conducted no more than one year before the AEDT is used
• Statistical analysis: The audit must calculate the selection rate and impact ratio for each category (race/ethnicity and sex/gender), testing for disparate impact
• Historical or test data: The audit should use historical data from the employer's own use. If unavailable, test data may be used with appropriate documentation
• Intersectional analysis: The DCWP final rules require analysis of intersectional categories where data is available (e.g., Hispanic male, Black female)

The impact ratio compares the selection rate of each category to the most selected category. While the law does not establish a specific threshold for acceptable impact ratios, the EEOC's four-fifths rule (0.80 ratio) is commonly used as a benchmark.

Areebi's compliance dashboards can track AEDT usage patterns and support audit preparation by providing comprehensive data on AI-assisted employment decisions. Audit trails document every AI interaction for auditor review.

Local Law 144 requires two forms of notice:

Candidate and Employee Notice

Employers must provide notice to candidates and employees who are subject to AEDT evaluation at least 10 business days before using the AEDT. The notice must include:

• That an AEDT will be used as part of the assessment or evaluation
• The job qualifications and characteristics the AEDT will assess
• Information about the data sources and data retention policy
• Instructions for requesting an alternative selection process or accommodation

Notice may be provided via the job posting, on the employer's website, or by direct communication to the candidate.

Public Posting of Bias Audit Results

Employers must publicly post a summary of the bias audit results on their website, including:

• The date of the most recent bias audit
• The selection rates and impact ratios for all tested categories
• The distribution of the AEDT's scoring or classification system

The summary must remain posted for at least six months after the AEDT is last used for an employment decision. Organizations can use Areebi's policy engine to enforce notification workflows and ensure compliance with timing requirements.

Local Law 144 is enforced by the NYC Department of Consumer and Worker Protection (DCWP). Penalties include:

• First violation: Civil penalty of $500
• Subsequent violations: Civil penalties between $500 and $1,500 per violation
• Each day of use of a non-audited AEDT constitutes a separate violation
• Failure to provide notice to a candidate or employee constitutes a separate violation

While individual penalty amounts may appear modest, the per-day and per-candidate accumulation means that organizations using AEDTs without compliance can face significant aggregate penalties. An employer using a non-audited AEDT for 30 days with 100 candidates could theoretically face penalties exceeding $150,000.

Beyond DCWP enforcement, organizations should note that bias audit failures may also create exposure under federal anti-discrimination laws (Title VII, ADA) and state human rights statutes. Request a demo to see how Areebi helps organizations manage AEDT compliance risk.

Organizations using AI in hiring should take the following steps to comply with Local Law 144 and prepare for similar requirements in other jurisdictions:

1. Inventory AEDTs: Identify all AI tools used in hiring and promotion decisions. Include third-party vendors, embedded features in HRIS/ATS platforms, and shadow AI tools used by hiring managers.
2. Engage an independent auditor: Select a qualified, independent auditor to conduct the bias audit. Ensure the auditor has experience with statistical analysis of employment data and AEDT evaluation.
3. Compile audit data: Gather historical data on AEDT usage, including selection rates by race/ethnicity and sex/gender. Use Areebi's audit trails to compile comprehensive interaction data.
4. Conduct the bias audit: Complete the audit and review results for potential disparate impact. Address any identified issues before continuing AEDT use.
5. Implement notice processes: Configure automated candidate notifications at least 10 business days before AEDT use. Post audit summaries on your careers website.
6. Establish ongoing governance: Use Areebi's compliance dashboards to monitor AEDT usage and policy engine to enforce governance controls continuously.

Explore our pricing plans to implement AI governance for HR and talent management. Visit the Trust Center for detailed compliance documentation.