EEOC, DOJ Release Expectations on Employers’ Use of Technology, AI for Employment Decisions
For decades, employers have used technology to help decision-making, from hiring to performance bonuses. While seemingly taking human biases out of the equation, the U.S. Equal Employment Opportunity Commission (EEOC) and the Department of Justice (DOJ) have voiced concerns over potential disability discrimination from the use of technology.
To help employers identify and avoid the potential pitfalls of using decision-making software, including artificial intelligence (AI), on May 12, 2022, the EEOC released a technical assistance document (TAD) about technology and disability discrimination. The EEOC’s TAD is part of its Artificial Intelligence and Algorithmic Fairness Initiative “to ensure that the use of software, including artificial intelligence (AI), machine learning, and other emerging technologies used in hiring and other employment decisions comply with the federal civil rights laws that the EEOC enforces.”
The DOJ also, on May 12, released a guidance that focuses on algorithms and AI in the hiring process.
While they do not create new or different legal obligations for employers, these documents provide insight on how these agencies view AI at work.
The EEOC’s TAD applies the Americans with Disabilities Act (ADA), including regulations and existing guidance, where technology intersects with workplace legal issues. The TAD addresses traditional recruiting and hiring tools, such as personality and cognitive assessments, as well as cutting edge technologies such as AI.
The TAD focuses on the three “most common ways that an employer’s use of algorithmic decision-making tools could violate the ADA.” According to the EEOC, these are:
The employer does not provide a reasonable accommodation for an employee or applicant with a disability to be rated “fairly and accurately”;
The tool intentionally or unintentionally “screens out” individuals with disabilities; and
The tool violates the ADA’s restrictions on disability-related inquiries and medical examinations.
The TAD also provides what the EEOC calls “promising practices” for employers that may minimize the risk of claims, including:
Training staff to recognize and process requests for reasonable accommodations as quickly as possible.
Training staff to develop or obtain alternative means of rating job applicants and employees when the current evaluation process is inaccessible or otherwise unfairly disadvantages applicants or employees who have requested a reasonable accommodation because of a disability.
Working with outside parties who administer decision-making tools to ensure proper accommodations are provided.
Using algorithmic decision-making tools designed to be accessible to individuals with as many kinds of disabilities as possible.
Informing all job applicants and employees who are being rated that reasonable accommodations are available for individuals with disabilities.
Providing clear and accessible instructions for applicants and employees to request accommodations.
Describing in plain language and accessible formats the traits the tool is designed to assess, the assessment methods used, and the variables or factors that may affect the rating.
Ensuring that the algorithmic decision-making tools measure only abilities or qualifications that are “truly necessary” for the job, even for people who have on-the-job reasonable accommodations.
Ensuring that necessary abilities or qualifications are measured directly, rather than indirectly.
Confirm with vendors that tools do not ask questions likely to elicit information about a disability or seek information about an individual’s physical or mental impairments or health, unless the inquiries relate to a request for reasonable accommodation.
While some of the “promising practices” are already utilized by employers, some of them arguably go beyond compliance and seek changes in practice that are not without potential drawbacks. For example, many companies do not detail the traits an assessment is measuring or the variables or factors that affect the rating. In addition, some assessments measure characteristics that are correlated with abilities and qualifications because it may be difficult or impossible to objectively measure abilities of applicants directly. Again, this technical assistance document does not have the force of law.
The DOJ guidance focuses on the hiring process, discusses how use of these technologies may screen out people with disabilities, and how to avoid potential discrimination. The DOJ recommends that employers test technologies they intend to use and consider reasonable accommodations they can provide.
Employers should also be thoughtful of other potential discrimination claims. As the EEOC points out, employers and software vendors should take steps to ensure that algorithmic decision-making tools are appropriately vetted to prevent disparate impact discrimination under Title VII of the Civil Rights Act based on race, sex, national origin, color, or religion. But those steps are generally different from the ones needed to ensure compliance with the ADA given that the unique nature of every disability requires individual consideration. Decision-making tools that have been “validated” for some purposes, for example, still may inappropriately screen out an individual because of a disability.
Employers also should be mindful of data privacy issues arising from the collection of information necessary to apply the algorithms. In some states, employers may wish to discuss with counsel issues of notice, consent, acceptable use, disclosure, reasonable safeguards, and retention of this information.
The EEOC and DOJ are not alone in worrying about the use of AI and related technology in workplaces. In 2019, Illinois passed the Artificial Intelligence Video Interview Act (AIVI Act), which imposes consent, transparency, and data destruction requirements on employers that implement AI technology during the job interview process. The first state law to regulate AI use in video interviews, the AIVI Act took effect January 1, 2020. In 2020, Maryland likewise enacted a law that requires notice and consent before use of facial recognition technology during a job interview. In 2021, New York City Council passed a measure that creates certain obligations for employers that use AI in hiring practices. Finally, more is on the horizon. California’s Fair Employment and Housing Council is considering draft regulations regarding automated-decision systems in the workplace.
Technology-based decision-making tools are more accessible than ever, particularly as new technologies and applications emerge to drive businesses’ response to marketplace demands and to serve the growing number of remote and hybrid workplaces. From sports and athletic organizations trying to improve athlete performance, to logistic companies striving to improve driver safety records, to driving efficiencies in the food service business, organizations in nearly all industries are wise to consider a myriad of legal issues potentially affecting every facet of recruiting and human capital management.