The AI Joy Paradox: Satisfied Workers, Rising Cognitive Load

New BCG research reveals something counterintuitive: the same workers who say AI has boosted their job satisfaction are also the ones reporting increased cognitive load. It's not a contradiction — it's the AI joy paradox, and most HR teams are only measuring one side of it. The data is striking. Among regular AI users, 67% say AI has improved their satisfaction at work. But 41% of that same group report higher cognitive load since they started using it. The reason? Nearly half of workers now spend more time managing and directing AI than doing their actual work. AI hasn't eliminated the burden — it's shifted it from doing to directing. The retention risk is hiding in that gap. Most HR teams track engagement and satisfaction. Very few measure cognitive load as a distinct signal. Gallup's 2026 findings add fuel to the fire: employees with active manager support for AI are 8.7 times more likely to say their work has genuinely transformed. The difference isn't the technology — it's whether someone helps workers integrate it without absorbing all the coordination cost themselves. In this episode, we break down the paradox, what's driving it, and the three things HR needs to do: measure cognitive load alongside satisfaction, build managers who coach AI workflows, and redesign the processes where AI overhead is highest. BCG's future-built companies are already doing this — and they're 5x more likely to be doing strategic workforce planning as a result.