Employers who rely on artificial intelligence driven tools for their recruiting and hiring processes may face new regulations in New York and New Jersey. In the past few weeks, three bills have been proposed (two in New Jersey and one in New York) that follow New York City’s Local Law 144 (“Local Law 144”), a law that sets rules for certain automated employment decision tools used by employers in New York City, which became effective last July. Similar bills were introduced in both states during the 2023 legislative session.
A bill (A. 9314) proposed in the New York Assembly on February 28, 2024, would impose requirements on employers who use artificial employment decision tools (“AEDTs”) in hiring processes. This bill is similar to Local Law 144, which we reported on here and here. However, the new Assembly bill only targets tools used for screening and hiring applicants, while Local Law 144 also included tools used for assessing employees for advancement. The bill would add a new section to the New York Labor Law, which would make it an unlawful employment practice for employers to screen applicants with an AEDT for jobs “within the state,” unless the AEDT was subject to a disparate impact analysis in the past year. The bill defines AEDT as “any system used to filter employment candidates or prospective candidates for hire in a way that establishes a preferred candidate or candidates without relying on candidate-specific assessment by individual decision-makers.” While the bill explicitly states that the disparate impact analysis would not need to be publicly filed and would be “subject to all applicable privileges,” it does require that prior to the implementation of an AEDT, employers would need to post a “summary of the most recent disparate impact analysis” and “the distribution date of the tool” on its website. The employer would also be required to provide the state’s Department of Labor with this summary on an annual basis. Unlike Local Law 144, New York’s proposed law does not provide for civil monetary penalties for employer violations. Rather, the state’s Attorney General or Department of Labor Commissioner would be able to initiate investigation into possible violations and could bring actions in court to correct alleged violations. If enacted, the bill would go into immediate effect.
Similarly, the New Jersey legislature introduced two bills (A. 3854 and A. 3911) which also aim to regulate employers’ use of AI tools in hiring. A. 3854, like New York’s proposed bill, generally mirrors Local Law 144 and seeks to regulate AEDTs in hiring. The stated goal of the bill is to “minimize employment discrimination that may result from the use of the tools.” A. 3854 first defines an AEDT as “any system the function of which is governed by statistical theory, or systems the parameters of which are defined by systems…which automatically filter candidates or prospective candidates for hire or for any term, condition, or privilege of employment in a way that establishes a preferred candidate or candidates.” Unlike the New York bill, A. 3854 would make it unlawful to sell, or offer to sell, an AEDT in the state unless (i) the AEDT has been the subject of a bias audit conducted in the past year prior to its sale; (ii) the sale of the AEDT includes, at no additional cost, an annual bias audit service that provides the results of such audit to the buyer; and (iii) the AEDT is sold, or offered for sale, with a notice stating that the AEDT is subject to the provisions of the proposed law. Moreover, any “person” who uses an AEDT to screen a prospective job candidate would have to provide a summary on its website of the most recent bias audit, and notify each job candidate within 30 days of the AEDT’s use that an AEDT was used in connection with their candidacy, identifies the job qualifications or characteristics that the AEDT used to assess the candidate and, upon the candidate’s request, identifies the source of the data collected and the employer’s data retention policy. Penalties for violations would range from $500 for a first offense (and each violation occurring on the same day as the first one) and a $500 to $1500 penalty for each subsequent offense. Finally, the bill specifies that violations accrue on a daily basis for each AEDT that is sold or offered for sale in violation of the proposed law, and that “each instance” in which notice is not provided to a job candidate within 30 days constitutes a single violation, while each 30-day period thereafter constitutes a separate violation. If the bill is enacted, it would take effect on the “first day of the third month next following enactment.”
The second New Jersey bill, A. 3911, concerns the use of AI-enabled video interviews during the hiring process. The bill provides that “an employer in the State” that requests job candidates to record video interviews, and uses AI analysis of the videos submitted, would be required do the following: (i) “notify the applicant before the interview that [AI] may be used to analyze the applicant’s video interview and consider the applicant’s fitness for the position;” (2) “provide an applicant with information before the interview explaining how the [AI] works and what general types of characteristics it uses to evaluate applicants;” and (3) “obtain, before the interview, written consent, which may be electronic, from the applicant to be evaluated by the [AI] program.” The proposed law would prohibit employers from using AI analysis to evaluate applicants who have not consented to its use. Further, within 30 days of receiving a request from an applicant, an employer would have to delete the applicant’s interviews and instruct any other persons who received copies to delete the videos. The proposed law would also direct employers who use AI analysis of video interviews to “collect and report” the race and ethnicity of applicants who are and are not offered opportunities for in-person interviews, or who are and are not offered positions. This information would need to be reported on an annual basis to the state’s Department of Labor and Workforce Development (the “Department”). The Department would then report to the Governor and State Legislature whether the data discloses racial bias in the use of AI. Penalties for violations would include a $500 fine for a first offense and a $1,000 fine for subsequent offenses. If enacted, the proposed law would take effect immediately.
We expect that we will see AI related legislative and regulatory activity this year in multiple jurisdictions. We will keep monitoring further developments.