“Workslop” may not be a familiar term, but it seems to be a familiar experience for many American office employees.
Researchers from BetterUp Labs and Stanford over 40% of desk workers have received so-called in the last month.
It’s costing workers time and potentially costing companies millions in lost productivity.
What’s more, they say workslop is a byproduct of newly deployed generative artificial intelligence tools, which are ironically intended to make companies more productive.
“It’s AI-generated content masquerading as good work that doesn’t actually advance the work itself,” said one of the researchers, , a communication professor and the founding director of the Stanford Social Media Lab.
They said sloppy or nonsensical AI-generated content – such as emails, reports or presentation slides – can pass the burden of work from the creator to the recipient.
And that exerts a toll on workers and companies, they said.
“It came with an emotional cost. They were often annoyed, or angry or confused,” Hancock said. “It came with a bit of a relational cost. They said they were less likely to trust the person in the future. So, it has this important trust piece. And time and effort. They had to decide, ‘Do I say something to this person? Do I just redo it myself?'”
Employees reported spending close to two hours on average dealing with each instance of workslop from a colleague.
The researchers estimated that could lead to over $9 million a year in lost productivity for an organization with 10,000 workers.
The researchers pointed to and reports showing both a striking rise in AI use at work and a striking lack of measurable returns on investment for the companies deploying AI tools for their employees to use.
The researchers believe the workslop phenomenon might be one reason why companies aren’t getting the intended bang for their buck from AI initiatives.
Are companies mandating AI use without clear guidance?
Indiscriminate mandates might play a role, according to researchers.
But Hancock also said some employees might just be using AI “to check the box” and get rid of what they see as less-important busy work.
, an AI expert and the co-director of the Johns Hopkins Institute for Assured Autonomy, said there are really good uses for generative AI in the workplace.
But he said the practice is in a transition phase, and it’s possible employees are unwittingly using AI in a way that’s actually creating more work for their colleagues.
“This technology is early,” Dahbura said. “I think ultimately there will be guidelines, norms and even etiquette that will emerge.”
Hancock said their research into AI-generated workslop started this summer, but it’s an observational offshoot of something he called “AI slop” on the internet.
AI slop, he said, is low-quality and mostly worthless AI-generated content seemingly created just to drive web traffic.
“And we started to notice a similar kind of thing where AI-generated content was in the workplace,” he said.
The crucial difference is that workslop occurs between people, so it carries more risk of damaging co-worker relationships.
They said workslop is usually sent between peers (40%), but it is also sent to and from managers.
Their study participants expressed feeling annoyed (53%), confused (38%) and offended (22%) when they received workslop from someone in their office.