Q: What artificial intelligence concerns should employers consider in hiring?

A: Title VII of the Civil Rights Act of 1964 (Title VII), the Age Discrimination and Employment Act (ADEA), and the Americans with Disabilities Act (ADA) all provide various protections for employees against discrimination based on a number of protected classes. Although artificial intelligence (AI) seemingly would eliminate bias by putting the decision-making authority in the “hands” of a computer instead of an individual, the reality appears to be that AI is often riddled with bias, implicit or explicit, or discriminatory screening processes. In employment, facially neutral algorithms are being used more frequently to screen applicants. But in certain circumstances, these algorithms disproportionately disadvantage certain protected classes.

AI lends itself to particularly significant concerns with respect to individuals with disabilities. AI decision-making tools could potentially screen out applicants because of a disability despite implementing facially neutral algorithms. For example, the program may have an algorithm designed to screen out all applicants who have significant gaps in their employment history. If an individual had a gap in employment due to disability, such as undergoing prolonged treatment, then the algorithm would screen out that person due to their disability in violation of the ADA. Other issues may arise if an algorithm is designed to screen out applicants based on a lack of certain skills that could be remedied through a reasonable accommodation.

Disparate impact also is crucial to the discrimination analysis as it recognizes that discrimination is not always intentional and facially neutral policies can have a disproportionate negative impact on protected classes. For example, a stringent criminal background screening process through AI – without taking into consideration numerous factors or exceptions – could disproportionately affect certain races over others due to historical and systematic racism related to arrests and incarceration. Under the disparate impact analysis, employers are prohibited from implementing these policies and practices.

Technical Assistance Documents

Although there are currently no federal regulations governing the use of AI in employment, in May 2023 the Department of Justice (DOJ) and Equal Employment Opportunity Commission (EEOC) each released technical assistance documents on disability discrimination to guide employers when using AI to make employment decisions – the first official guidance from any agency on AI considerations in hiring. The DOJ’s guidance, Algorithms, Artificial Intelligence, and Disability Discrimination in Hiring, and the EEOC’s guidance, The Americans with Disabilities Act and the Use of Software, Algorithms, and Artificial Intelligence to Assess Job Applicants and Employees, both highlight issues employers should consider to ensure that hiring algorithms do not discriminate against applicants with disabilities. The documents highlight practices that reduce the likelihood of discrimination.

Caveat

Employers should be aware that their hiring policies are not exempt from the requirements of Title VII, the ADA or the ADEA, based simply on the fact that they are screened by an algorithm by a third-party vendor. The use of algorithms for screening potential employees must be narrowly tailored for only necessary purposes and carefully reviewed for bias and potential disparate treatment. Wilson Elser’s Employment & Labor Practice is equipped to do this for our clients.