New York City Issues Final Rules for Automated Employment Decision Tools Law, Postpones Enforcement Again

On April 6, 2023, the New York City Department of Consumer and Worker Protection (“DCWP”) issued final rules covering the City’s Automated Employment Decision Tools (“AEDT”) Law that was enacted on December 11, 2021. With the issuance of the final rules, the DCWP also announced that enforcement of the law will be further delayed from April 15, 2023, to July 5, 2023, providing covered employers with additional time to comply with the recently announced requirements.

Summary

The AEDT law prohibits New York City employers and employment agencies from using AEDTs in their employment decision-making processes, unless the employer complies with several affirmative requirements before using any such tool. These preconditions include (i) conducting an annual bias audit; (ii) making the results of any bias audit(s) publicly available on the employer’s or employment agency’s website; and (iii) providing at least ten days’ notice to employees and job candidates that an AEDT will be used in connection with a specific position or assessment, and providing details concerning how an applicant or employee can request an alternative selection process (if one is available) or an accommodation under existing law.

Definitions

The final rules modify the definition of “automated employment decision tool” to expand its scope beyond the definition provided in prior versions of the proposed rules. The AEDT law defines the term “automated employment decision tool” as “any computational process, derived from machine learning, statistical modeling, data analytics, or artificial intelligence, that issues simplified output, including a score, classification, or recommendation, that is used to substantially assist or replace discretionary decision-making for making employment decisions that impact natural persons.”

The DCWP’s final rules modify the AEDT definition by expanding the definition of the term “machine learning, statistical modeling, data analytics, or artificial intelligence” beyond its prior meaning. The final rules define this term to mean “a group of mathematical, computer based techniques: (i) that generate a prediction, meaning an expected outcome for an observation, such as an assessment of a candidate’s fit or likelihood of success, or that generate a classification, meaning an assignment of an observation to a group, such as categorizations based on skill sets or aptitude; and (ii) for which a computer at least in part identifies the inputs, the relative importance placed on those inputs, and, if applicable, other parameters for the models in order to improve the accuracy of the prediction or classification.” This definition constitutes an expansion of the term “machine learning, statistical modeling, data analytics, or artificial intelligence” due to DCWP’s elimination of the requirement, from an earlier version of the proposed rules, that such techniques use inputs and parameters that “are refined through cross-validation or by using training and testing data.”

On a positive note for covered employers, the final rules have retained the narrow definition of the “substantially assist or replace” prong of the AEDT definition, which applies only where the AEDT is utilized to (i) rely solely on a simplified output (such as a score, classification or ranking); (ii) use a simplified output as one of a set of criteria where the simplified output is weighted more heavily than any other criterion in the set; or (iii) use a simplified output to overrule conclusions derived from other factors including human decision-making. Practically speaking, this means that a decision tool may only be covered under the new law if its output is the most significant or only factor used in screening a candidate for employment or employee for a promotion.

Bias Audit

Under NYC’s AEDT law, covered employers or employment agencies may use an AEDT as long as the tool has been subject to a compliant bias audit by an “independent auditor” no more than one year prior to the use of the tool. The bias audit requirement is intended to evaluate the disparate impact an AEDT has on a group of job applicants or employees based on sex, race, and ethnicity—the same demographic characteristics that mirror EEO-1 reportable data.

The final rules clarify that, to satisfy the requirements of the new law, a bias audit must meet the following minimum requirements:

1. Calculate the “selection rate” for each category.
2. Calculate the “impact ratio” for each category.
3. Ensure that the two “selection rate” and “impact ratio” calculations separately calculate the impact of the tool on (i) sex categories; (ii) race/ethnicity categories; and (iii) intersectional categories of sex, ethnicity, and race (e.g., impact ratio for selection of Hispanic or Latino male candidates vs. Not Hispanic or Latino Black or African American female candidates).
4. Ensure the above calculations are performed for each group, if an AEDT classifies candidates or employees into specified groups (e.g., leadership styles).
5. Indicate the number of individuals the AEDT assessed that are not included in the required calculations because they fall within an unknown category.

The DCWP’s final rules further clarify that the independent bias audit may exclude from the calculations of the impact ratio a demographic category that represents less than 2% of the data being used for the bias audit and that, where any such category is excluded, the summary of the results must include a justification from the independent auditor regarding the exclusion, as well as the number of applicants and scoring or selection rate for the excluded category.

Under the new law, an independent bias auditor is defined to mean “a person or group that is capable of exercising objective and impartial judgment on all issues within the scope of a bias audit of an AEDT.” According to the final rules, an auditor cannot be considered an “independent auditor” for purposes of the law if the auditor (i) is or was involved in using, developing, or distributing the AEDT; (ii) at any point during the bias audit, has an employment relationship with an employer or employment agency that seeks to use or continue to use the AEDT or a vendor that developed or distributes the AEDT; or (iii) at any point during the bias audit, has a direct financial interest or a material indirect financial interest in an employer or employment agency that seeks to use or continue to use the AEDT or a vendor that developed or distributed the AEDT.

Use of Historical Data

The final rules allow employers to use historical data and test data in bias audits under appropriate circumstances. Multiple employers or employment agencies who use the same AEDT may rely on the same bias audit conducted using historical data of other employers or employment agencies, but only under the limited circumstances where the employer or employment agency provided historical data from its own use of the AEDT to the independent auditor conducting the bias audit or where the employer or employment agency has never used the AEDT. The final rules also authorize employers and employment agencies to use test data in a bias audit if insufficient historical data is available to conduct a statistically significant audit. If test data is used, the audit must explain why historical data was not used and describe how the test data was generated and obtained.

Publication and Notice Requirements

The final rules require employers and employment agencies to post the date of the most recent bias audit, with a summary of the results and the date that they began using the AEDT, on the careers or job section of their websites in a clear and conspicuous manner. The summary and distribution date must be posted for at least six months after the latest use of the AEDT for an employment decision. Employers can meet these requirements with “an active hyperlink to a website containing the required summary of results and distribution date, provided that the link is clearly identified as a link to results of the bias audit.”

NYC’s AEDT law and final rules also require that employers provide at least ten days’ notice to candidates and employees who reside in New York City that an AEDT will be used in connection with a given assignment or decision. Employers can provide the required notice via any of the following methods: (i) on the employment section of an employer’s website; (ii) in a job posting; or (iii) U.S. mail or email. Such notice must also inform employees and applicants how to request an alternative selection process or a reasonable accommodation under other applicable laws, if available. Importantly, however, the final rule clarifies that employers are not required to provide an alternative selection process.

Conclusion

The enforcement date for the AEDT law and final rules has been delayed until July 5, 2023. However, employers and employment agencies in New York City should act now to determine whether they or any employment agencies with whom they contract are covered under the new law and, if so, begin to take the steps necessary to comply with the requirements of the AEDT law and final rules beginning July 5. Employers who use a tool that is or may be covered should evaluate whether the costs of compliance with the AEDT law and final rules exceed the benefits of continued use of a covered AEDT. Employers should also consider whether it is necessary or advisable to work with third-party vendors to ensure compliance with applicable legal requirements, and to determine whether any updates are needed to existing contracts to account for the risks and burdens imposed by the new law.

If you have any questions about New York City’s AEDT law or final rules, please contact Charles H. Kaplan (646.218.7513), Glen P. Doherty (518.433.2433), Kinsey A. O'Brien (716.848.1287), Andrew D. Drilling (716.848.1344), or any other member of our Labor & Employment Practice.

Disclaimer:

This Client Alert is a form of attorney advertising. Hodgson Russ LLP provides this information as a service to its clients and other readers for educational purposes only. Nothing in this Client Alert should be construed as, or relied upon, as legal advice or as creating a lawyer-client relationship.