The AI Atrophy Problem: How CIOs Fight It
Executive Take
Leaders should mandate documented AI-verification checkpoints and "show your work" protocols now, before dependency on unchecked AI outputs becomes embedded in team workflows and harder to reverse.
Executive Summary
At MIT Sloan's 2026 CIO Symposium, leaders discussed "AI atrophy" — the erosion of critical thinking as AI use grows. Michael Schrage advised treating AI outputs as testable hypotheses, not answers. George Westerman suggested verifying AI tools fit the task. Other strategies: protecting unstructured thinking time, building verification checkpoints, and requiring teams to show prompts and edits used.
Why It Matters
AI and HR leaders should care because organizations optimizing for speed via AI adoption risk quietly degrading employee judgment and critical-thinking capacity, a workforce capability risk that compounds over time and is hard to measure until it's acute.