The AI Fluency Gap

Leaders think almost nobody has adopted AI seriously. They're wrong by a factor of three. McKinsey's Superagency in the Workplace survey asked executives how many staff use AI for at least 30% of their daily work. Executives guessed 4%. Employees reported the real number: 13%. The gap is visibility, not confidence. Leadership is planning policy for a workforce that already moved on without telling them.

Fluency breaks into five separate skills, and most people have only picked up one or two of them by accident. Fundamentals: knowing what the model can and can't do. Interaction: giving it context it can actually use. Application: building it into a real workflow, not a party trick. Strategy: knowing which problems are worth pointing it at. Growth: keeping up as the tools change under you. The AI Fluency Assessment scores all five separately. A high score in one doesn't cover for a gap in another.

Usage and confidence used to move together. In 2026, they split. ManpowerGroup's Global Talent Barometer 2026 found regular AI use jumped 13 points to 45% of workers this year. In the same survey, confidence in using the technology fell 18%. People are doing more with less certainty they're doing it right. More usage won't fix this gap; it's what's exposing it.

Nobody chooses to stay unfluent. The loop just doesn't leave an exit.

The payoff for closing the gap isn't abstract. PwC's 2025 Global AI Jobs Barometer found AI-skilled roles carry a 56% wage premium, up from 25% the year before. Pew found 75% of AI-knowledgeable workers feel positive about their own job prospects, against a workforce where roughly half say they're worried about AI's impact on their future. Three moves close the gap.