EEOC Issues Guidance on AI Use in Employment Decisions
On May 18, 2023, the U.S. Equal Employment Opportunity Commission (EEOC) issued new guidance on use of artificial intelligence in hiring, transfer, demotion, monitoring performance, determining pay or performance and other employment selection procedures titled, “Assessing Adverse Impact in Software, Algorithms, and Artificial Intelligence Used in Employment Selection Procedures Under Title VII of the Civil Rights Act of 1964.” In the guidance, the EEOC outlined considerations for incorporating automated systems into these employment decisions.
The document did not address intentional discrimination, but instead focused solely on disparate impact discrimination cases. Disparate impact discrimination occurs when a practice of an employer or policy is facially neutral but unduly and usually unintentionally impacts a protected group, e.g. testing all applicants and using the results of the test to eliminate disproportionately a group of minorities. Many employers do not even realize the disparate effects of a practice or policy until they are shown statistics in a lawsuit of how the practice or policy has impacted a minority group.
The EEOC defined three very important terms in trying to address the issue: software, algorithm, and artificial intelligence (AI). Software was defined as “information technology programs or procedures that provide instructions to a computer on how to perform a given task or function.” Algorithm means a “set of instructions that can be followed by a computer to accomplish some end.” AI was defined as a “machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.”
The guidance focused on the possible discriminatory effect that algorithmic decision-making tools may have in the workplace. Examples given of when employers may possibly use AI included:
- Resume scanners that use keywords to prioritize applications and resumes
- Monitoring software that monitors an employee’s keystrokes and measures or rates employees based on the results
- Utilizing chatbots that ask candidates about their qualifications and automatically disqualify candidates that do not satisfy pre-defined characteristics or qualifications
- Using interviewing software that filters out candidates based upon their facial expressions or speech patterns
- Testing software that provides a score for each candidate regarding their personality or cognitive skills
The EEOC warned that without proper safeguards, the use of these AI tools may disproportionately have a disparate impact on certain minority groups. While the EEOC guidance is not binding, employers and their counsel would be wise to recognize the import of this document. The EEOC is simply placing employers on notice that automatic and blind application of these AI tools, may lead to disparate discrimination. After issuing the guidance,
EEOC Chair Charlotte Burrows said, “[a]s employers increasingly turn to AI and other automated systems, they must ensure that the use of these technologies aligns with the civil rights laws and our national values of fairness, justice, and equality.” She believes the guidance will “aid employers and tech developers as they design and adopt new technologies.”
In other words, employers have been placed on notice that any software, algorithm, or AI tool that they use for screening and that they use for any employment decision may produce disparately discriminatory practices. The EEOC recommended that employers frequently audit their systems to ensure the technologies are not violating pre-existing laws on disparate impact. They also suggested oversight of any vendor that uses these same tools to ensure the technologies are not disproportionately affecting minority groups.