The most enthusiastic adopters of AI in the workplace are now showing the first clear signs of burnout—despite the widespread promise that AI would make work easier, not harder. The dominant narrative in American work culture has long been that AI won’t destroy jobs, but rather liberate workers from tedious tasks, allowing them to focus on higher-value work and ultimately work less. This message has resonated deeply with white-collar professionals eager to embrace change and avoid obsolescence. The idea was simple: AI as a force multiplier. Lawyers, writers, analysts, and coders could do more with less effort, become more valuable, and reclaim time for themselves. But a new study published in the Harvard Business Review, based on eight months of immersive research at a 200-person tech company, reveals a different reality. The researchers found that when employees genuinely adopted AI tools—without being pushed or pressured—their workloads didn’t shrink. Instead, they expanded to fill the time AI had created. Employees began working through lunch breaks and into late evenings, not because they were forced to, but because they felt compelled to keep up with rising expectations. One engineer captured the shift perfectly: “You thought that because you could be more productive with AI, you’d save time, work less. But then you don’t work less. You just work the same or even more.” This pattern wasn’t isolated. In interviews with over 40 employees, researchers observed a consistent trend: the ability to do more led to doing more, not less. On Hacker News, a developer echoed the sentiment: “I feel this. Since my team adopted an AI-first workflow, expectations have tripled, stress has tripled, and actual productivity has only gone up by maybe 10%. It feels like leadership is pressuring everyone to prove AI is worth it, so we’re all working longer hours just to show we’re keeping up.” These findings aren’t entirely new. A previous study from last summer found that experienced developers using AI tools took 19% longer to complete tasks—even though they believed they were 20% faster. Meanwhile, a National Bureau of Economic Research analysis of AI adoption across thousands of workplaces found productivity gains averaged just 3% in time savings, with no meaningful changes in earnings or work hours across occupations. What sets this new research apart is its focus on real-world behavior, not just metrics. It doesn’t dispute that AI can enhance human performance—it confirms it. But it also shows where that enhancement leads: not to rest, but to fatigue, burnout, and an ever-tightening grip of work on personal time. As one researcher put it, “Organizational expectations for speed and responsiveness are rising, and employees feel increasingly unable to step away.” The tech industry’s bet has been that empowering people with AI would solve the problem of inefficiency. But what’s emerging is something more complex: a system where increased capability fuels greater demand, not relief. The result may not be a productivity revolution, but a new kind of work culture—one where the tools meant to free us are instead deepening our exhaustion.