Workers using AI more report increased frustration despite time savings, according to Glean study
Workers using artificial intelligence tools in the workplace are experiencing growing frustration, even as the technology saves them an average of 11 hours per week, according to a 2024 report by Glean’s Work AI Institute. The study found that 87% of digital workers use AI, but 6.4 hours per week are lost to “botsitting”—a term describing the labor of feeding AI context, supervising outputs, and debugging errors, said Rebecca Hinds, head of the Work AI Institute.
AI adoption outpaces integration, creating friction
The report highlights a disconnect between AI’s potential and its current implementation. While 87% of workers use AI tools, only 13% believe the technology has significantly improved their company’s performance. Hinds noted that employees often juggle multiple disconnected AI systems, exacerbating the workload. “We’re seeing high rates of multiple tool usage, and often those tools aren’t connected,” she said.
A 2023 McKinsey & Company analysis found similar challenges, noting that 60% of employees using AI tools reported “frustration with inconsistent or unreliable outputs.” The Glean study adds that workers spend 6.4 hours weekly on tasks like cleaning up AI-generated work, a figure corroborated by a 2024 Stanford University survey of over 2,000 knowledge workers.
Productivity gains offset by hidden labor costs
The study calculates that AI automates 27% of workers’ tasks, saving 11 hours weekly. However, these savings are largely offset by the time spent managing AI systems. A 2024 Harvard Business Review article on AI deployment in enterprises noted that “the hidden labor of AI integration often outweighs its efficiency gains,” a sentiment echoed by Glean’s findings.
Companies are beginning to address the issue. Microsoft, for example, announced in April 2024 a new AI integration platform aimed at reducing tool-switching overhead. “The goal is to create a unified workflow where AI tools communicate seamlessly,” said a company spokesperson. Glean’s report suggests such efforts could mitigate the “vicious cycle” of AI adoption described by Hinds.
What’s next for AI in the workplace?
Experts warn that without better integration, worker frustration could hinder AI adoption. A 2024 Pew Research Center survey found that 58% of workers worry AI will increase their administrative burdens. Hinds argues that companies must invest in “AI literacy training and cohesive system design” to realize the technology’s full potential.
The European Union’s AI Act, which takes effect in 2026, includes provisions for “human-AI collaboration standards,” signaling regulatory attention to these challenges. As AI becomes more embedded in work, the balance between efficiency and usability will define its success.