How Manufacturers Can Avoid Potential Pitfalls of AI-Based Recruitment

The manufacturing industry is familiar with strategically integrating automation tools and technologies to address labor shortages and promote safety and productivity in the workplace. Manufacturers exploring the potential uses of artificial intelligence (AI) in recruitment as a natural extension of their automation efforts should consider the legal concerns — including potential discrimination and bias inherent in such use.

AI in manufacturing may be able to measure cognitive impairment to ensure safe and proper equipment use as with fatigue monitoring and other assessments, thereby overcoming implicit biases in the process when a manager performs the same assessment. That said, not all AI systems are created equal, and AI itself is not inherently unbiased.

With or without AI, hiring decisions that discriminate based on race, color, religion, sex, national origin, age, sexual orientation or gender identity, physical or mental disability, military status, genetic information, and any other protected status are prohibited under law. Executive Order 14110, issued on Oct. 30, 2023, recognized the inherent potential for unlawful discrimination when AI is used in the hiring process. Office of Federal Contract Compliance Programs (OFCCP) guidance for federal contractors highlights that AI is only as unbiased as the data it learns from and may thus perpetuate inequalities already in the hiring process and workplace. Equal Employment Opportunity Commission (EEOC) guidance for employers on use of AI in employment selection procedures notes that employers can evaluate “an adverse impact on a particular protected group by checking whether use of the procedure causes a selection rate for individuals in the group that is ‘substantially’ less than the selection rate for individuals in another group.”

Using the seemingly benign parameter of significant time gaps in the applicant’s employment history to screen out resumes, for example, may result in automatic disqualification of applicants who took time off for the birth of a child — having a disproportionate impact on female applicants. The outcome (screening out otherwise qualified female applicants) may be directly attributable to underdeveloped or improperly implemented AI systems, but manufacturers using those systems can be liable for any resulting discrimination or bias.

Similarly, as the nature of manufacturing operations often requires a 24/7 work environment and commensurate work schedules, part of the hiring process may consider applicants’ availability, including on nights and weekends, for shift work. To avoid potential discrimination claims, AI screening has to be calibrated to avoid automatically disqualifying applicants whose unavailability on weekends is attributable to religious observances, including attendance at religious services.

While the OFCCP guidance pertains directly to federal contractors and the EEOC guidance is considered technical assistance, all manufacturers should note the concerns raised on the use and integrity of AI in the hiring process. Manufacturers must carefully consider and evaluate the AI being used. They also should:

• Inquire about the steps the vendor took (such as assessments and studies of system bias) to determine whether use of the AI may cause a substantially lower selection rate for applicants within a legally protected category 
• Analyze and articulate the job-relatedness of the hiring selection procedure 
• Periodically review the use of AI to assess any inherent bias or discrimination

Finally, manufacturers must consider state and local laws concerning AI in the hiring process, including in Illinois, Maryland, and New York City. Illinois expanded its AI-related requirements, amending the Illinois Human Rights Act to prohibit discriminatory use of AI (intentional or not) in recruitment, hiring, and all other aspects of the employment relationship effective Jan. 1, 2026.

(Law Graduate Madelyn Foster contributed to this article.)