Last week, amid much anticipation, the EEOC released a new technical assistance document for employers, providing guidance on the use of artificial intelligence (AI), while ensuring compliance with Title VII. Entitled “Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures Under Title VII of the Civil Rights Act of 1964,” the guidance addresses AI used in employment selection procedures under Title VII, and follows the prior EEOC’s guidance on the use of AI tools and the Americans with Disabilities Act from May 2022.

As technology advances, the use of automated systems in the workplace has become increasingly common. Employers often rely on AI to assist in decision-making at different stages of the employment process. Examples of such software include resume scanners, virtual assistance to ask job candidates about their qualifications, and video interviewing software.  Historically, employers have monitored their traditional decision-making processes to make sure that they comply with Title VII and do not disproportionately affect certain protected groups.  Up until now, employers have been left to their own devices (literally and figuratively) to determine how to monitor the impact of artificial intelligence systems.

So, what can be done? The EEOC suggests that employers can assess whether a selection procedure has an adverse impact on a particular protected group “by checking whether use of the procedure causes a selection rate for individuals in the group that is ‘substantially’ less than the selection rate for individuals in another group.” To assess adverse impact, employers can use the “four-fifths rule,” which is a general rule of thumb for determining the selection rate for applicants. If the selection rate for a protected group is less than 80% of the selection rate of another group, it may indicate potential discrimination.  However, employers must not rely solely on the four-fifths rule.  The EEOC recognizes that it is not always appropriate, “especially where it is not a reasonable substitute for a test of statistical significance.”

Employers should consider evaluating their AI tools for potential risk, even if the test was developed by a third party, outside vendor. If use of that tool has an adverse impact on individuals based on race, color, religion, sex, or national origin, or for a combination of such characters, then it may violate Title VII unless the employer can show that such use is “job related and consistent with business necessity.” 

The world of artificial intelligence and algorithmic decision continues to evolve.  We will be monitoring any additional guidance from the EEOC relating to artificial intelligence, and any resulting enforcement actions.