Inadvertent Disability Discrimination May Lurk in Hiring Software, Artificial Intelligence and Algorithms
Inadvertent Disability Discrimination May Lurk in Hiring Software, Artificial Intelligence and Algorithms

While employers can realize significant time savings or improvement in the hiring process by implementing technology, employers must ensure that the selected technology is designed and implemented in a way that prevents discrimination based on protected classes such as disability, race, sex, national origin, color, or religion.  Employers may be responsible for violations of the Americans with Disabilities Act (ADA) and/or the Fair Employment and Housing Act (FEHA) even if the discrimination is not intentional or the technology was developed by an outside vendor.

The Equal Employment Opportunity Commission (EEOC) recently issued guidance on how to avoid disability discrimination claims regarding the use of hiring technology such as software, algorithms, and artificial intelligence (AI) to automate or augment stages of the employment process. (Under the ADA, disability is defined as a physical or mental impairment that, when left untreated, “substantially limits” one or more “major life activities.”) In particular, to avoid disability discrimination in the use of hiring technology an employer must 1) select technology and use it in a way that does not intentionally or unintentionally “screen out” an individual with a disability, even though that individual is able to do the job with a reasonable accommodation, 2) not allow the technology to ask impermissible disability-related inquiries and medical examinations, and 3) provide a reasonable accommodation that is necessary for a job applicant or employee to be rated fairly and accurately by the software, AI, or algorithm.

Hiring technology can assist in screening applicants by scoring their resumes, measuring their skills and abilities through computer based tests, evaluating whether an applicant meets job qualifications, holding online interviews of applicants, and using computer-based tests to measure an applicant’s skills or abilities.  Hiring technology accomplishes these feats by following set instructions to accomplish a particular objective.  For example, hiring technology may, in scoring applicant resumes, search for specific words in a group of resumes.  Another example would be the use of hiring technology during interviews to recognize an applicant’s facial expressions during a video interview.  While these tools may promote efficiency, technology is only as good as the set of instructions used to develop it, and developers must be sure that these instructions do not contain any bias.  Hiring technology must be both designed and used in a way that will avoid adverse impact or disparate impact based on protected classes, including adverse or disparate impact on individuals with disabilities.

To avoid disability discrimination, hiring technology must evaluate a candidate’s job skills without unlawfully screening out or penalizing candidates with disabilities, or making the application process harder for such individuals.  Individuals with disabilities may be unlawfully screened out in a number of ways.  For example, an applicant who has limited manual dexterity because of a disability may have difficulty completing a knowledge test that requires the use of a keyboard, trackpad, or other manual input device even though the applicant possesses the requisite knowledge.  As such, the knowledge test may not accurately assess the applicant’s knowledge as a result of its inability to account for the applicant’s disability.

Similarly, an applicant with a speech impediment may be affected by interviewing software that analyzes verbal responses to assess the ability to solve problems.  If the interviewing software is unable to identify and properly account for the applicant’s responses because of a speech impediment, it will not be able to accurately assess the ability to solve problems and, consequently, may penalize the applicant for the speech impediment.  Qualifying limitations include disabilities that make one or more major life activities “more difficult, painful, or time-consuming to perform as compared to the way that most people perform them.”  If the technology would screen an applicant out due to a disability, penalize an applicant for having a disability, or make it more difficult, painful, or time-consuming for someone with a disability to perform the task, then the employer may need to offer a reasonable accommodation.

Hiring technology may also violate the ADA if it poses “disability-related inquiries” or seeks information that qualifies as a “medical examination” before giving the candidate a conditional offer of employment.  These types of inquiries are prohibited even if the applicant does not actually have a disability.  Additionally, inquires may discriminate based on protected disabilities even if they do not make impermissible inquiries about a disability.  For example, an algorithm that screens out an applicant for answers consistent with bipolar disorder or depression can violate the law even though it does not ask the individual to disclose his or her disability.

The EEOC’s recent guidance, mentioned and linked above, provides insight into how these issues can be avoided.  In addition to carefully selecting hiring technology to avoid bias and impermissible inquiries, employers using hiring technology may avoid disability discrimination by informing applicants at the outset that reasonable accommodations are available to individuals with disabilities and providing directions for requesting reasonable accommodations.  Employers should make clear that a request for accommodations does not impact an applicant’s chance of being hired.  These notices can be provided directly by the employer, by the software in use, or by a vendor facilitating the use of the software.  Once a candidate informs an employer of a medical condition or disability that may impact the ability to be evaluated by hiring technology, employers must engage in the interactive process to determine whether the employee has a qualified disability and whether there is an alternative option that can be provided as a reasonable accommodation.  Reasonable accommodations are fact specific, but may include using a different input method, using a different format, or a more accurate assessment of the applicant’s skill.  If there is a reasonable accommodation, the employer must provide it unless doing so would involve an undue hardship.

Employers seeking to harness hiring technology should assess the extent to which the hiring technology might intentionally or unintentionally discriminate against protected classes, including individuals with disabilities.  Additionally, employers should assess their notice of the right to a reasonable accommodation and plans for responding to requests for accommodations relating to hiring technology before implementing it. 

For more information regarding any aspect of the ADA, including the practical considerations relating to the implementation of hiring technology, please contact the authors of this post or your legal counsel.

This AALRR post is intended for informational purposes only and should not be relied upon in reaching a conclusion in a particular area of law. Applicability of the legal principles discussed may differ substantially in individual situations. Receipt of this or any other AALRR publication does not create an attorney-client relationship. The Firm is not responsible for inadvertent errors that may occur in the publishing process. 

©2021 Atkinson, Andelson, Loya, Ruud & Romo

Other AALRR Blogs

Recent Posts

Popular Categories

Contributors

Archives

Back to Page

Necessary Cookies

Necessary cookies enable core functionality such as security, network management, and accessibility. You may disable these by changing your browser settings, but this may affect how the website functions.

Analytical Cookies

Analytical cookies help us improve our website by collecting and reporting information on its usage. We access and process information from these cookies at an aggregate level.