Why AI Doesn't Make Companies More Productive
Computerworld, Friday, August 28th, 2026
Computerworld's Mike Elgan says AI overload, not employee resistance, explains stalled productivity.
Mike Elgan notes that despite $2.59 trillion in projected 2026 global AI spending, US productivity grew just 0.07% and roughly 95% of enterprise generative-AI pilots show no measurable bottom-line impact.
He challenges a University of Pittsburgh study blaming worker sabotage for a resistance 'doom loop,' arguing it shows correlation, not causation, and that fearful workers actually have incentive to visibly use AI.
Instead he proposes 'AI overload': individual productivity gains dump verification and processing burden onto colleagues, citing doubled job applications per posting and legal AI filings rising from 1% to 18% since 2023, and calls for redesigning AI tools around organizational, not individual, output.