The Equal Employment Opportunity Commission (EEOC) has ramped-up enforcement and guidance in recent months over employers’ use of artificial intelligence (AI).
On May 18, 2023, as part of its Artificial Intelligence and Algorithmic Fairness Initiative, the EEOC issued its second technical assistance document (TAD) concerning AI, addressing employment selection procedures under Title VII of the Civil Rights Act of 1964. Just a couple of weeks earlier, on May 5, 2023, the EEOC filed suit against a group of related entities that offer tutoring services to students in China under the brand name “iTutorGroup.” Now, according to a consent decree notice of settlement filed in federal court on or about August 9, 2023, in what may be the EEOC’s first settlement of AI-related claims, the parties settled the suit for $365,000, along with a long list of corrective action items for the defendants.
The EEOC’s allegations are fairly straightforward: The defendants hire tutors from the United States to provide online English-language tutoring to adults and children in China. Each year, they hire thousands of tutors who submit applications through the defendant’s website. The defendant’s website requested the date of birth of applicants and programmed their application software to automatically reject female applicants age 55 and older and male applicants age 60 and older. In this case, the charging party’s initial application, which indicated she was older than 55, was rejected. She then resubmitted her application with a more recent date of birth as the only change. She was offered an interview. The defendants rejected more than 200 other presumably qualified applicants because of age. According to the EEOC, these employment practices violated Section 4 of the Age Discrimination in Employment Act, 26 U.S.C. Section 623(a) and (b).
In its TAD, the EEOC observed that employers sometimes rely on different types of software that incorporate algorithmic decision-making at a number of stages of the employment process. The examples it provides include resume scanners that prioritize applications using certain keywords; or “virtual assistants” or “chatbots” that ask job candidates about their qualifications (perhaps even age) and reject those who do not meet pre-defined requirements. These algorithmic decision-making tools can have the effect of unlawfully screening out otherwise qualified candidates.
In this case, the sophistication of the application software and the nature of the AI that allegedly screened out applicants of a certain age were unclear. What is clear is the EEOC, through its Artificial Intelligence and Algorithmic Fairness Initiative, is increasing its efforts to ensure the use of software (including AI, machine learning, and other emerging technologies used in hiring and other employment decisions) comply with the federal civil rights laws the agency enforces.
No doubt, algorithmic decision-making tools, or “automated employment decision tools” as they are referred to under the New York City AI law, can significantly boost the productivity and results of an organization’s recruiting efforts. However, the development and implementation of those tools also come with significant compliance and litigation risk. Issues organizations rolling out these tools should consider include:
- Understanding the use case, e.g., recruiting, performance monitoring, performance improvement, and so on.
- Tracking the application and requirements of emerging laws, guidance, and established frameworks.
- Considering application of guardrails or key principles, such as notice, informed consent, transparency, privacy and security, fairness, nondiscrimination, and ability to understand and challenge outcome.
- Incorporating “promising practices” suggested by the EEOC, such as in connection with ensuring reasonable accommodations are available.
- Oversight of the use of the tool from procurement through implementation.
- Vetting the vendor and the product offering the tool.
- Record retention obligations.
