New York City Passes Landmark AI Bias Law, with Similar Legislation Across the Country Likely on the Rise

New York City will soon enact novel legislation, Local Law Int. No. 1894-A, aimed at curbing bias in hiring and promotion decisions aided by artificial intelligence (“AI”), which has increasingly been employed by businesses in recent years.  The new law, which requires employers to conduct an independent audit of the automated tools they use, marks the first time employers in the U.S. will face heightened legal requirements if they wish to use AI decision-making tools.  While the new law was passed late last year, it is set to take effect throughout New York City in January 2023—leaving employers scrambling and job applicants wondering how exactly the new law protects them.

The use of AI in hiring and employment decisions—such as résumé screening, automated video interviews, and assessing reasonable accommodation requests by employees with disabilities—has been on the radar of the U.S. Equal Employment Opportunity Commission (“EEOC”) and lawmakers across the country for some time.  AI tools can range from algorithms built to find ideal candidates to software that assesses body language, and they have faced scrutiny in recent years for their potential to perpetuate bias against protected groups.  The basic premise of these concerns is simple: humans are capable of implicit bias; therefore, so are the AI tools that humans program.

Lawmakers’ and regulators’ interest in addressing AI bias was initially sparked by Charges of Discrimination filed with the EEOC alleging AI-related workplace discrimination.  As a result, the EEOC recently issued guidance explaining how reliance on AI to serve functions typically reserved for human resources (no pun intended) can result in disability discrimination, and has the EEOC has indicated that it will continue to delve further into the issue.  Further, Illinois and Maryland have already passed legislation related to AI bias, though the scope of those laws is considerably narrower than New York City’s (with Maryland’s law being especially narrow, as it merely required applicants’ consent prior to the use of facial recognition technology in job interviews).  Most recently, an AI bias bill similar to New York City’s was introduced in Washington, D.C., and the California Civil Rights Department announced that it is drafting regulations to bring automated decision-making tools expressly under the purview of existing employment discrimination laws.  AI bias in employment is even on the radar of the U.S. Senate, where the Algorithmic Justice and Online Platform Transparency Act was introduced just last year.  

While these laws and regulations variously fall short of addressing the problem or are still works in progress, Local Law Int. No. 1984-A appears to have teeth.  The new law’s key terms are as follows:

•    New York City employers will be prohibited from using automated employment decision tools to screen job applicants, unless the technology has been subject to an independent bias audit to assess potential disparate impact at least one year prior.
•    Companies will be required to notify employees and applicants if AI was used to make job decisions.
•    In addition to requiring independent audits, the law will require employers to make the results of those audits publicly available on their websites. 
•    Employers will also be required to disclose the data their AI tools collect, whether by publishing it proactively or in response to an inquiry.
•    Before using AI tools to assess job applicants residing in New York City, employers will be obligated to provide at least ten days’ notice regarding, among other things, accommodations that may be available.

Employers who violate the new law will be subject to fines of $500 for initial violations, and $1,500 for each subsequent violation.  This means that new $1,500 fines will accrue daily if the unlawful conduct is not corrected. 

Notably, the law does not include a private right of action; however, there is obvious potential for class action litigation if, for example, the City (or another state or municipality that passes similar legislation down the line) finds an AI tool used by an employer to be discriminatory.

While myriad questions regarding the passage, application, and enforcement of these laws remain, there is no question that regulating AI bias in employment decisions is on the rise—and rightfully so—with New York City currently leading the charge.