AI productivity paradox: Workers lose 37% of time savings fixing AI mistakes and low-quality output

AI Speeds Up Work While Creating Hidden Costs

Artificial intelligence promises to transform workplace efficiency, yet the AI productivity paradox reveals a more complicated reality. While 85% of employees report saving one to seven hours per week using AI tools, roughly 37% of that time vanishes into correcting low-quality AI output, rewriting content, and verifying results 1

1

3

3

. This phenomenon, increasingly referred to as AI workslop, represents what Workday researchers call an "AI tax on productivity"—for every 10 hours of efficiency gained through AI, nearly four hours are lost to fixing its output 4

4

.

The data comes from a November survey by Workday of 3,200 employees across North America, Europe, and Asia at companies with at least $100 million in revenue and 150 employees 3

3

. The findings expose a critical gap between AI's promise and its current performance, particularly in areas like data analysis and visualizations (55%), research and fact-finding (52%), long-form reporting (52%), and writing and marketing content (44-46%) 2

2

.

The Cost of AI Errors Extends Beyond Time

Only 14% of employees consistently achieve net-positive outcomes from AI use, according to the research 1

1

3

3

. The burden falls heaviest on the most frequent AI users, who spend disproportionate time verifying and correcting output. Highly engaged employees lose an average of approximately 1.5 weeks per year to rework 4

4

.

Research from Zapier adds that three in five workers (58%) spend more than three hours per week revising outputs, with more than one-third (35%) spending over five hours and 11% spending over 10 hours weekly tidying up generated content 2

2

. Despite 92% agreeing AI improves overall productivity, only 2% say AI outputs need no revision 2

2

.

The cost of AI errors extends beyond wasted hours. Many organizations have experienced rejected work (28%), security incidents or privacy breaches (27%), customer complaints (25%), and compliance or legal issues (24%) 2

2

. "There is a big productivity paradox," Gerrit Kazmaier, president of product at Workday, told Axios. The most frequent users of AI invest the most time in reviewing and correcting what it produces 3

3

.

AI Training Gaps Undermine Productivity Gains

The root cause isn't just technology—it's the lack of investment in people. Nearly 9 in 10 organizations report that fewer than half of their roles have been updated to include AI-related skills 4

4

. AI has been "layered onto roles that were never updated to accommodate it," forcing employees to reconcile faster output with unchanged expectations around accuracy and accountability 4

4

.

The impact of AI training is dramatic. Nearly all (94%) of trained workers say AI boosts productivity, but only 69% of untrained workers agree. Only 1% of trained workers report productivity declines 2

2

. Yet a disconnect exists between leadership priorities and employee experience—while 66% of leaders cite skills training as a top investment priority, only 37% of employees most exposed to AI-related rework report increased access to training 4

4

.

Emily Mabie, Senior AI Automation Engineer at Zapier, explained: "The companies seeing the best results aren't the ones avoiding AI. They're the ones who have invested in training, context, and orchestration tools that turn AI from a sloppy experiment into a managed process" 2

2

.

Upskilling Employees for AI Requires Strategic Investment

Organizations realizing the greatest value from AI "treat saved time as a strategic resource," reinvesting in upskilling their teams, improving collaboration, and strengthening judgment and creativity-driven work 1

1

. Experts recommend making AI training compulsory for all workers handling AI tools, prioritizing high-risk teams and tasks first 2

2

.

Companies can support workforce development by providing prompt templates and formalizing review processes 2

2

. Job redesign becomes essential—roles, skills, and processes must catch up with new ways of working brought on by AI 1

1

. Most employees (77%) review AI-generated work just as carefully as work done by humans, if not more 1

1

.

Workday's report emphasizes that "paying a high tax on AI efficiency is not inevitable. It is the cost of implementing AI without investing in the humans who use it" 4

4

. The solution, according to Mabie, isn't fewer tools—it's better infrastructure 2

2

. As AI tools rapidly improve at performing real-world work, the challenge for organizations is ensuring their people can effectively harness these capabilities while minimizing the burden of refining outputs from AI tools.