Across offices and Slack channels everywhere, a quiet paradox is emerging.
Generative AI tools are being rolled out with fanfare, hailed as the next great leap in workplace productivity.
Employees are using them enthusiastically, leaders are investing heavily – and yet, for most organisations, the payoff simply isn’t there.
AI use at work has nearly doubled since 2023, and the number of companies running fully AI-led processes has doubled too. Yet a recent MIT Media Lab report found that 95 percent of organisations see no measurable return on their AI investments.
The problem isn’t exactly with the technology itself – it’s with how people are using it.
When Good Work Looks Good, But Isn’t
In a joint study between BetterUp Labs and the Stanford Social Media Lab, researchers identified a phenomenon spreading quietly through offices, AI workslop.
Borrowed from "AI slop" – a phrase for low-quality, machine-generated content flooding social media – workslop refers to AI-created output that looks like quality work but lacks the depth, accuracy, or context to actually move a project forward.
It’s the report that’s beautifully formatted but factually thin, or the summary that sounds smart but misses the point.
“AI has made sloppy work faster,” Jamie Aitken, VP of HR Transformation at Betterworks told UC Today.
“It’s not that AI creates the problem – it just accelerates it.”
The result is a new kind of burden: the illusion of productivity.
The creator feels efficient, but the receiver inherits the confusion. Someone has to decode, correct, or redo what the machine produced, shifting the cognitive load downstream.
The Workslop Tax
BetterUp’s research suggests the problem is far from rare.
In a recent survey of more than a thousand full-time US employees, 40 percent reported receiving workslop in the past month.
Those incidents cost an average of one hour and 56 minutes each – or roughly $186 per employee per month in wasted time – nearly $9 million a year for a 10,000-person organisation.
But the hidden costs go beyond the balance sheet – receiving AI-generated low-effort work leaves more than half of employees feeling annoyed, confused, or even offended.
Forty-two percent say they view the sender as less trustworthy afterward, while a third say they’d rather not work with that person again.
“Low-effort AI work isn’t just inefficient – it’s corrosive,” Aitken says. “It breaks down trust and collaboration, which are exactly what you need for AI to work well.”
Over time, these small acts of artificial efficiency can quietly undermine the human connections that make work function.
Each instance of workslop carries an invisible tax – not just in time, but in relationships.
Pilots and Passengers
Behind these behaviours lies a deeper divide in mindset. BetterUp’s researchers describe two archetypes emerging in the age of AI: pilots and passengers.
Pilots use AI purposefully – to explore ideas, sharpen creativity, and accelerate meaningful progress.
