A recent survey of managers of U.S. businesses points to a rapidly shifting environment in which key employment decisions — including hiring, promotion and termination — are increasingly influenced and in some cases driven by artificial intelligence (AI).
According to the survey by ResumeBuilder of 1,342 business leaders with hiring and firing authority, 60 percent said they use AI tools to make decisions about their direct reports. Still higher percentages of these managers said they have used AI to determine raises (78 percent), promotions (77 percent), layoffs (66 percent) and even terminations (64 percent).
The survey indicates AI is not only used at the hiring stage but is now deeply integrated throughout the employee lifecycle. Managers reported varied uses of AI, such as sifting applications, tracking employee productivity and engagement, screening for skill gaps and flagging employees at risk for turnover.
The survey also shows a high level of trust accorded to AI’s judgment. Approximately 46 percent of respondents believe AI is better than a human at making objective decisions about promotions, while 45 percent said they rely on AI recommendations on employee terminations. Perhaps more striking is that one in five managers said they often let AI make employment decisions without human review or input. Proponents of AI use point to its objectivity and ability to synthesize massive amounts of data quickly, features that are particularly attractive for large organizations managing thousands of employees.
However, this reliance on AI for critical employment decisions raises significant legal concerns. Federal equal employment opportunity laws, such as Title VII of the Civil Rights Act, the Americans with Disabilities Act (ADA) and the Age Discrimination in Employment Act (ADEA), prohibit discrimination based on protected characteristics including race, color, sex, religion, national origin, disability, and age. An AI system that disadvantages employees or job candidates from protected groups — intentionally or not — could trigger disparate impact liability under these laws.
AI algorithms are only as unbiased as the data and models on which they are built. If historical data used to train AI systems reflect prior biases or systemic inequalities, those biases may be perpetuated or even exacerbated by AI-driven processes. An AI tool trained on a company’s previous promotion outcomes may ‘learn’ to favor employees from majority groups that were overrepresented in past promotions, regardless of qualifications. Similarly, AI productivity assessment could disfavor workers with disabilities or older workers whose work styles deviate from data-driven norms embedded in the system.
In light of these findings, employers must proceed with caution when integrating AI into employment decisions. With assistance of experienced employment and labor counsel, companies should adopt best practices for subjecting AI tools to rigorous bias auditing and for ensuring that human judgment remains central in all significant personnel actions. Failing to do so may not only expose organizations to legal liability but also risk undermining workplace diversity, inclusion, and trust.
The Reddy Law Firm, P.C., in Alpharetta advises Georgia employers on compliance with federal and state laws governing the workplace. Call 678-629-3246 or contact us online to schedule a free consultation.
