Employers are increasingly using AI-assisted tools to screen resumes, rank applicants, evaluate video interviews, and identify candidates who appear to be the “best fit” for a role. While these technologies may improve efficiency, they also create significant legal risks that many employers may not fully anticipate.
One of the biggest concerns is that AI hiring tools can unintentionally perpetuate discrimination. Many AI systems are trained using historical hiring data, meaning the technology learns patterns from an employer’s past hiring decisions. If those historical decisions reflect unconscious bias or disproportionate exclusion of certain groups, the AI system may replicate those same outcomes. To address these concerns, the Equal Employment Opportunity Commission (“EEOC”) created the Artificial Intelligence and Algorithmic Fairness Initiative, which examines how employers’ use of AI and other automated systems may implicate federal anti-discrimination laws and provides guidance regarding the lawful use of those technologies in the workplace. Information on this initiative can be found here.
Importantly, the EEOC has taken the position that employers may still face liability under federal anti-discrimination laws even when hiring decisions rely on AI-assisted tools developed or administered by third-party vendors. The EEOC’s fact sheet on this can be found here. The EEOC has warned that employers may violate Title VII where AI-assisted hiring tools disproportionately screen out applicants based on protected characteristics such as race, sex, or national origin, even absent any express intent to discriminate. Given that these technologies are often used to evaluate large numbers of applicants simultaneously, discriminatory outcomes may occur on a broader scale and may be more difficult for employers to detect if the underlying systems are not properly monitored or audited.
The EEOC has also made clear that monitoring AI-assisted hiring is an enforcement priority. In EEOC v. iTutorGroup, Inc., the EEOC alleged that an automated hiring system rejected older applicants based on age by programming its software to automatically reject female applicants age 55 or older and male applicants age 60 or older. The matter ultimately settled for $365,000. Additional details about this case can be found here.
More recently, a case in California, Mobley v. Workday, Inc., has drawn national attention. In that case, the plaintiff alleged that Workday’s AI-powered applicant screening tools discriminated against applicants based on race, age, and disability in violation of federal anti-discrimination laws. More about this case and the roadmap of its litigation can be found here. This case is significant because it raises questions about whether software vendors themselves may face liability. The important takeaway: employers cannot simply rely on third-party AI vendors without independently evaluating whether those systems create discriminatory hiring outcomes or disproportionately screen out protected groups.
The ADA also presents unique risks in the AI hiring space. According to guidance issued by the U.S. Department of Justice, AI-driven technologies may unlawfully screen out qualified individuals with disabilities if they are not properly designed or monitored. For example, software that evaluates speech patterns, facial expressions, eye contact, or response timing during video interviews may disadvantage applicants with speech impairments, neurological conditions, anxiety disorders, or other disabilities. Likewise, timed assessments or computer-based testing may violate the ADA if employers fail to provide reasonable accommodations. The DOJ further emphasized that employers are still responsible for ensuring their hiring technologies comply with the ADA, regardless of whether the systems are administered internally or through outside vendors. Details of the DOJ’s guidance on this issue can be found here.
Although there is potential risk, AI is here to stay, and employers must adapt or be left behind. So, what are some simple things employers do to mitigate against this risk in using AI-assisted technologies in hiring?
- Carefully vet AI vendors and examine any contracts with AI vendors regarding indemnification, waivers, insurance provisions, and apportionment of liability.
- Ask questions and understand how AI hiring tools are trained and evaluated.
- Regularly test for adverse impact on protected groups.
- Maintain meaningful human oversight regarding the hiring process.
- Ensure that applicants have a way to request accommodations during the hiring process.
The Employment Attorneys at Cavitch, Familo & Durkin are closely monitoring the use of AI in the employment lifecycle and are hosting an AI In The Workplace Panel discussion to further this conversation. This event will bring together legal and business professionals to discuss the practical realities of implementing AI in the workplace, the evolving legal and ethical considerations surrounding its use, and what employers can do now to prepare for the future of work. To attend the event, sign up here.