AI isn't closing the skill gap—it's turning it into a cliff.
April 17, 2026
Original Paper
Delegation and Verification Under AI
SSRN · 6463045
The Takeaway
We've been told AI is the great equalizer, but it’s actually creating a 'phase transition' where tiny differences in human ability lead to massive differences in outcomes. The study found that workers who are slightly worse at verifying AI output start over-delegating tasks, which slowly degrades the quality of the entire institution. Even if everyone's 'baseline' success goes up, the gap between those who can check the AI and those who just trust it becomes a structural divide. It’s not just about who uses AI; it’s about who has the power to double-check it. This means the future workplace won't just be about 'using' AI, but about avoiding the trap of becoming its passenger.
From the abstract
As AI systems enter institutional workflows, workers must decide whether to delegate task execution to AI and how much effort to invest in verifying AI outputs, while institutions evaluate workers using outcome-based standards that may misalign with workers' private costs. We model delegation and verification as the solution to a rational worker's optimization problem, and define worker quality by evaluating an institution-centered utility (distinct from the worker's objective) at the resulting