AI and algorithmic systems are increasingly being used by companies to set employee compensation, a phenomenon dubbed "algorithmic wage discrimination" or "surveillance pay." This practice, initially observed in app-controlled ride-hail and food-delivery sectors, is now expanding to traditional industries like healthcare, customer service, logistics, and retail. These systems leverage extensive real-time data, often collected through automated monitoring, to establish compensation brackets and calculate individual wages.
A first-of-its-kind audit of 500 AI labor-management vendors revealed that many employers are purchasing tools that feed worker data into automated pay decisions. These tools can analyze personal information, including credit history, social media activity, and past pay acceptances, to estimate the lowest salary a candidate might accept. This approach, which focuses on perceived desperation rather than qualifications or market rates, has drawn criticism from researchers and labor advocates. Some vendors even offer platforms capable of real-time pay adjustments and the steering of pay tiers and bonus structures, often with minimal human oversight.
Unlike traditional contractually negotiated wages, those determined by machine-learning systems are variable, uncertain, and opaque to workers, leading to increased stress, higher workplace injury rates, and decreased job satisfaction. A vast majority of the reviewed products lack transparency or feedback mechanisms, giving workers no visibility into the data or logic behind their pay calculations. Furthermore, many vendors apply identical performance benchmarks across diverse roles, disregarding crucial factors like task complexity, local market conditions, and worker accommodations.
While some companies, like Colgate-Palmolive and Intuit, deny using such tools for compensation decisions, the trend suggests a fundamental shift in the relationship between work and pay. Although HR teams are cautiously experimenting with AI in compensation, concerns remain regarding compliance with federal laws like the Fair Labor Standards Act and equal employment opportunity laws, as AI trained on historical data could amplify existing pay disparities. Experts emphasize that human review and approval of AI outputs remain crucial to ensure fairness and compliance.
