You open a document a colleague sent. It looks finished. Headings, bullet points, a confident tone, the right length. Then you start reading and realise it says almost nothing. It is fluent and empty, and now you have to do the actual thinking it only pretended to do.
That is workslop, and there is now a name and a price tag for it.
What exactly is workslop?
Researchers at BetterUp Labs and Stanford’s Social Media Lab defined it in 2025 as AI-generated content that masquerades as good work but lacks the substance to meaningfully advance a task.1 Its defining feature is not that it is low quality. It is that it shifts the burden of the work downstream, from the person who created it to the person who receives it.2 The sender looks productive. The receiver inherits the real job.
How much is it costing?
More than you would guess. In a survey of about 1,150 US full-time workers, roughly forty percent had received workslop in the previous month. Each instance took, on average, nearly two hours to deal with, at an estimated cost near 186 dollars per worker per month.3 Multiply that across a large organisation and the “time savings” from AI start to look like an accounting trick: time saved by the sender, time lost by everyone downstream, often at a net loss.
Does it damage more than time?
Yes, and this is the part that should give pause. Recipients reported lower trust in, and lower regard for, colleagues who sent them workslop, and reduced willingness to work with them again.4 The polished empty document does not just cost two hours. It quietly corrodes your professional reputation, because people can tell when they have been handed a machine’s first draft dressed up as your finished thinking.
Why is it suddenly everywhere?
Because producing plausible-looking output became nearly free. For most of history, generating a long, well-formatted document took real effort, which acted as a natural filter on pointless work. AI removed the filter. Now the path of least resistance runs straight toward generating more, longer, more confident artefacts, whether or not anyone needed them.
How do you stop making it?
Two moves. First, eliminate before you automate. If a report is barely read, the answer is not to have AI generate it in four minutes and feel clever for saving thirty-six. That makes the pointless report permanent and longer, and pushes the cost onto whoever has to skim it. Delete it instead.
Second, do not abdicate. Around eighty-six percent of AI users already say they treat output as a starting point, not a final answer, and that is the correct instinct.5 Direct the model like a capable new hire, then verify and edit before anything goes out under your name. The test is simple: if you would be embarrassed to have written it yourself, do not send it. Workslop is what happens when speed is mistaken for value. The fix is to remember that the point was never to produce more. It was to produce something worth someone else’s time.
Frequently asked questions
What is workslop?
Workslop is AI-generated content that looks polished and complete but lacks the substance to actually advance the task, so it shifts the real work downstream onto whoever receives it. The term was coined in 2025 research from BetterUp Labs and Stanford.
How much does workslop cost?
In a survey of about 1,150 US workers, roughly 40 percent had received workslop in the previous month, each instance taking on average nearly two hours to address, an estimated cost near 186 dollars per worker per month.
How do you avoid creating workslop?
Do not hand over a whole task and look away. Treat AI output as a starting point, verify and edit it before it goes out under your name, and delete work that should not exist rather than generating more of it.
About the author
Tom Goodwin is the author of Don’t Work Harder, a book about taking the time AI gives back as time rather than more work. He is a co-founder of GAMEPLAN and writes on productivity, technology, and the economics of the working week.
Footnotes
Footnotes
-
Kate Niederhoffer, Gabriella Rosen Kellerman, et al. (BetterUp Labs and Stanford Social Media Lab), “AI-Generated ‘Workslop’ Is Destroying Productivity,” Harvard Business Review (September 2025). ↩
-
Niederhoffer et al. (2025): the defining feature is that workslop shifts the burden of the work downstream, from creator to receiver. ↩
-
Niederhoffer et al. (2025): in a survey of about 1,150 US full-time workers, roughly 40 percent received workslop in the previous month, each instance taking about one hour and 56 minutes to address, near 186 dollars per worker per month. ↩
-
Niederhoffer et al. (2025): recipients reported lower trust in and lower regard for colleagues who sent workslop. ↩
-
Microsoft Work Trend Index (2026): 86 percent of AI users say they treat AI output as a starting point, not a final answer. ↩