As companies rush to integrate AI tools into daily operations, a new survey from ResumeTemplates.com reveals that 31% of workers say their workload has actually increased since AI adoption, with some reporting they’re now expected to do twice, or even four times, as much work as before.
1 in 5 Workers Say They Can’t Keep Up without AI
AI has quickly become a workplace necessity rather than a nice-to-have tool. The survey found that about 1 in 5 workers (22%) say it would be difficult (17%) or impossible (5%) to keep up with their current workload without AI assistance. An additional 32% say it would be at least somewhat challenging to manage without these tools, meaning fewer than half of workers (46%) feel confident they could maintain their current pace without AI support.
Workers Are Now Expected to Do Twice As Much Work
Despite promises of increased efficiency, AI adoption hasn’t lightened the load for most workers. Nearly half of all workers (48%) now describe their workload as very or extremely heavy, and 31% say their workload has increased since AI was introduced at their company. Only 16% of workers report their workload has actually decreased.
“AI can cut down the busywork and help people focus on what really matters, but if professionals feel busier right now, it may be because they are,” says Julia Toothacre, chief career strategist at ResumeTemplates. “After years of layoffs and reorganizations, many teams are operating with fewer people, so remaining employees are carrying more. AI can help manage that load, but employees should watch for workload creep, when managers assume AI means you can take on more without providing the resources to support it.”
The pressure to use AI is explicit in many workplaces. Nearly half of workers (45%) say managers explicitly reference using AI when assigning additional work, making it clear that productivity expectations have been adjusted upward with the assumption that AI tools will help employees meet them. Attitudes toward AI are mixed overall, with 18% of workers saying they dislike (12%) or hate (6%) AI, while 36% like it and 17% love it.
“Organizations need to provide proper training and clear expectations about how and when AI should be used,” says Toothacre. “Some employees will pick it up quickly and thrive, but others will take longer and may not like using the new tools. It’s important to start small and show how AI can improve day-to-day efficiency, then build up to more complex use cases over time. If you force employees to use something without introducing it appropriately, you risk eroding trust and hurting morale. It’s not worth it when there are better ways to roll it out.”
Higher Expectations
For workers whose workload has increased, the expectations have become demanding.
- 33% say they are expected to do twice as much work as before.
- 6% say they are expected to do three times as much work.
- 4% say they are expected to do at least four times as much work.
- 57% say they are expected to do more work, but not double.
AI use isn’t just encouraged, it’s often expected or required.
- 21% of workers say AI use is required.
- 39% of workers say AI use is expected but not required.
Workers Experience AI Fatigue
- 37% say they have experienced AI fatigue, a state of exhaustion and overwhelm related to the demands of learning and using AI in their daily work.
- 40% say they are worried AI could replace them within the next five years, adding another layer of stress to an already challenging situation.
- 14% of workers say AI has decreased their job satisfaction.
Methodology
ResumeTemplates.com surveyed 1,250 full-time U.S. workers in corporate jobs who spend a significant amount of time working on a computer. The survey was conducted in February 2026 using Pollfish’s online survey platform. Pollfish uses random device engagement to reach respondents organically, ensuring a diverse and representative sample. The platform employs quality control measures that include attention checks, response time monitoring and duplicate detection to maintain data integrity. The margin of error is ±2.8% at a 95% confidence level.
