Artificial intelligence (AI) is set to transform how work is organized in the knowledge economy, according to a study by IESE Business School. The authors explore how AI changes the roles of workers, the size and structure of firms, and the distribution of benefits from new technologies.

The study distinguishes between two kinds of AI. Autonomous AI can work on its own, taking over tasks traditionally done by people. Basic versions push humans into more specialized roles, shrinking firms and lowering productivity. Advanced autonomous AI, on the other hand, shifts people into routine tasks, leading to bigger, more productive companies. While autonomous AI boosts overall output, it primarily rewards highly skilled workers, widening inequality.

“Basic autonomous AI displaces humans towards specialized problem solving, leading to smaller, less productive, and less decentralized firms,” the researchers explain. “In contrast, advanced autonomous AI reallocates humans to routine work, resulting in larger, more productive, and more decentralized firms.”

Levelling the playing field

Non-autonomous AI, which acts only as a helper, levels the playing field by giving less-skilled workers tools to solve tougher problems. This type of AI spreads the benefits more evenly but doesn’t increase productivity as much as autonomous AI.

The authors argue that policymakers face a trade-off: maximizing efficiency or reducing inequality. Their framework helps explain why studies show mixed results—some finding that AI benefits lower-skilled workers, others showing it favors the highly skilled. As AI advances, understanding these dynamics will be crucial to managing its economic impact.

“Our findings reveal a trade-off between output and inequality that policymakers encounter when regulating the autonomy of AI,” the researchers conclude.?