The appeal of artificial intelligence (AI) is more real than ever as we see an explosion of interest in using AI technology to make everyday life — and work — easier. Many retailers, for example, are considering AI to streamline their recruitment processes. While AI can conserve resources and accelerate hiring for high-turnover positions, employers need to consider the legal risks.

In May, the U.S. Equal Employment Opportunity Commission (EEOC) released a statement assessing the adverse impact of software, algorithms, and AI use in employment selection procedures. According to the EEOC, the use of AI to make predictions, recommendations, or decisions in the employment context may violate Title VII of the Civil Rights Act, even when applying neutral criteria, if it has an adverse effect based on race, color, religion, sex, or national origin. This means that employers can be on the hook for decisions made by an AI, even where they were not involved in the development of the selection process criteria or the creation of the algorithmic decision-making program itself.

Any employers considering AI must implement proper safeguards to ensure the system’s algorithms comply with civil rights laws. Self-analysis is critical, and employers should acquire an understanding of how their processes work, such as the purpose of the algorithm, the features assessed, and how those features are used to make decisions. Employers also must consistently monitor their data to identify whether the AI is causing any statistically significant disparities in selection rates by race, gender, national origin, and so on.

In addition, employers need to consider the specific locations at which they might implement AI hiring practices, as many states and localities are beginning to regulate the use of AI in employment selection procedures. New York City Local Law 144 (NYC 144) went into effect July 5, 2023, and the regulation requires employers to look at their use of automated employment decision tools (AEDTs), or AI, in making employment decisions such as hiring and promotions for candidates in New York City. Further, if an employer uses AEDTs, NYC 144 requires the employer to perform an independent bias audit within one year of the use of the AEDTs and make a summary of the audit publicly available. Additionally, NYC 144 requires an employer using AEDTs to provide applicants with notice of the tool’s use and implementation, as well as allow applicants to request an alternative selection process or an accommodation.

While companies understandably rely on the use of AI to speed up the application process, the use of such devices can lead to obstacles for employers if they fail to follow federal, state, and local regulations. For New York City, failure to follow NYC 144 can result in daily fines of between $500 and $1,500 per violation. Outside of financial penalties, the use of AI can cause unintended consequences, such as a lack of workforce diversity or disproportionate hiring. Although the idea is that AI will select successful candidates based on traits identified in application materials, a concern is that the new technology will identify an immaterial trait as material, leading to disproportionate hiring as to a protected class.