New York City "Regulate AI Use in Corporate Hiring and Promotion... Must Disclose Potential Discrimination"

The city of New York in the United States has become the first to regulate the use of artificial intelligence (AI) in companies' hiring and promotion decisions. The key points require companies to notify job seekers or employees about the use of AI software and to annually audit and disclose the potential for racial and gender discrimination. This is the first law in the U.S. to regulate AI use in hiring decisions.

[Image source=Reuters Yonhap News]

According to the Wall Street Journal (WSJ) and others, New York City will enforce the 'NYC 144' law starting on the 5th (local time). Passed by the New York City Council in 2021 and after about two years of public consultation, this law was created to reduce the possibility of biased decisions or potential discrimination in hiring and promotion processes by companies using AI software such as chatbot interview tools and resume scanners.

Accordingly, employers must notify job seekers and employees of the use of such AI software within a specified period in advance. Additionally, employers must annually inspect the software used in hiring and promotion decisions to determine whether there is any disadvantageous evaluation or discrimination against specific races or genders, known as the 'disparate impact' ratio, and disclose the results. Violations can result in fines of up to $15 million per day per violation for employers.

Victoria Lipnic, a partner at Resolution Economics, a labor consulting firm and former member of the U.S. Equal Employment Opportunity Commission, said, "The assumption is that employers who check the disparate impact ratio can take some form of action," adding, "It's quite a big assumption, but a first step."

Job seekers or employees cannot sue companies solely based on the disparate impact ratio. However, it can be used as potential evidence in various discrimination lawsuits filed under local and federal laws. The ratio is measured between 0 and 1, with values closer to 1 indicating little to no discrimination. WSJ explained, "If the disparate impact ratio for women is 0.3, it means that for every 10 male job applicants who pass the hiring process, only 3 female applicants do."

However, a low ratio close to 0 does not necessarily indicate discrimination. Whether the hiring criteria are job-related and consistent with business necessity plays a more important role. For example, WSJ mentioned that since Black and Latino individuals have lower college graduation rates than White and Asian Americans, if a college degree is a mandatory requirement for a position, companies can sufficiently defend themselves even if concerns arise about the low ratio for Black and Latino applicants.

In recent years, with the advancement of AI technology and the spread of online job applications, many companies have increasingly relied on AI software that can save time and costs in hiring and promotion decisions. However, concerns have also grown that AI software used by companies may amplify stereotypes and lead to biased decisions against minority groups. The recent surge in generative AI, such as ChatGPT, has further amplified these concerns. In the past, Amazon discontinued a pilot hiring algorithm program after realizing it excluded female college graduates.

Currently, similar bills have been proposed not only in New York City but also in Washington D.C., California, and Connecticut. Attorney Erin Cornell stated, "NYC 144 was designed to eradicate potential discrimination in employment decisions, but it is not an anti-discrimination law; it is a 'disclosure law,'" adding that the entire U.S. is watching New York as a pilot case for future technology regulation. WSJ evaluated, "Rather than regulating AI and software themselves, the focus is on mandating efforts to increase transparency."

This content was produced with the assistance of AI translation services.