Algorithm-Based Pay-for-Performance (APFP) systems: Paradoxes in artificial intelligence's influence on pay-for-performance theories

Although artificial intelligence (AI) and generative AI (GenAI) are increasingly used to assess and reward employees, their implications for foundational pay-for-performance (PFP) theories remain underexplored. Traditional PFP systems are effective in an era of static evaluations and infrequent feedback, but they lack the intelligence and flexibility needed for today's dynamic work environments. In response, we introduce algorithm-based PFP (APFP) systems—PFP systems that leverage AI and GenAI to enable real-time adaptability, predictive capabilities, customization, automated algorithmic recommending, and measurement sophistication. We then use the APFP framework to assess its implications for three foundational PFP theories (equity theory, expectancy theory, and tournament theory). The APFP framework integrates established PFP principles with AI and GenAI capabilities, reassessing how employees perceive, respond to, and engage with PFP systems. By conceptualizing how AI and GenAI influence the theoretical mechanisms of PFP, we offer a lens for understanding their influence on foundational PFP theories. Our theoretical contributions bridge existing PFP theories with emerging AI- and GenAI-driven environments to advance the literature and lay a foundation for future research that highlights inherent benefits and risks of APFP systems.