AI was supposed to make work cleaner. Instead, it’s making a mess.
Across US workplaces, a new term has entered the corporate dictionary — “workslop.” It refers to AI-generated work that “masquerades as good work, but lacks the substance to meaningfully advance a given task,” according to researchers at the Harvard Business Review (HBR).
The idea comes from a joint study by Stanford’s Social Media Lab and BetterUp Labs, which found that nearly 40% of full-time US employees had received AI-generated workslop in the past month. The study warns that these half-baked reports, slides, and emails are quietly eating away at productivity and workplace trust.
The hidden cost of ‘workslop’
According to HBR research, every time employees encounter workslop, they spend an average of one hour and 56 minutes trying to make sense of it or redo it entirely. That time adds up. For a company with 10,000 workers, the productivity loss could reach over $9 million per year.
Beyond lost time, researchers found that employees who receive workslop often feel:
- Annoyed (53%).
- Confused (38%).
- Offended (22%).
These emotional reactions are beginning to affect workplace relationships. Roughly 42% of respondents said they now view colleagues who send workslop as less trustworthy, while 32% said they’re less likely to work with those colleagues again.
A case study in AI gone wrong
In Australia, the consequences of AI misuse recently made headlines.
Bloomberg reported that consulting giant Deloitte Australia faced backlash after submitting a government report filled with “apparent AI-generated errors.” The $289,000 taxpayer-funded document reportedly misquoted a judge and included non-existent references.
Australian senator Barbara Pocock called the mistakes “the kinds of things that a first-year university student would be in deep trouble for.”
Deloitte later told media outlets that “the matter has been resolved directly with the client,” but the incident reignited global debates about AI’s limits and accountability in the workplace.
When AI creates more work instead of less
The irony, researchers say, is that employees are following orders.
Many companies have encouraged staff to integrate AI into their daily workflows. Yet, a McKinsey study shows that as many as 80% of companies using generative AI have seen “no significant bottom-line impact.” In some cases, 42% have even abandoned AI projects altogether.
Why? Because, as the Harvard team notes, some workers are using AI tools not to enhance their work but to avoid it, producing quick, impressive-looking content that’s hollow inside.
Who’s really to blame?
Columnist Gene Marks, writing for The Guardian, argues that while AI companies deserve scrutiny, the real responsibility lies with business leaders.
“In the workplace, the buck always stops with the boss,” he wrote, stressing that many employers deploy AI without proper training or guidelines.
Marks added that “in most cases, the software is not the problem. It’s the lack of investment in the people using it.”
Many companies are pushing AI tools without providing adequate training, clear policies on their use, or guidance on when AI is and isn’t appropriate. This leaves employees to figure it out on their own, leading to a “free-for-all” where the easiest path — copying and pasting AI output — becomes the norm.
The solution, researchers suggest, isn’t to abandon AI, but to implement it more thoughtfully. This involves training staff on how to leverage the tools and their limitations, establishing clear guidelines and reminding everyone that AI is a tool to augment human work, not replace human judgment.
After all, a polished report is only valuable if the information inside is solid.
In similar eWeek news: A report late last year found that nearly 7 in 10 job seekers say that enterprise AI fails to boost workplace performance.