Generative AI is producing a new category of office output that looks professional and reads as nothing. Researchers gave it a name - workslop - and put a number on the cost. The number is large.
Workslop: When Good Intentions Meet Bad AI Habits
In the world of corporate and white-collar jobs, a new term has taken hold: workslop.
🙋♀️Hey, what's workslop?
💡 Great question!
Workslop describes the epidemic of AI-generated outputs that look like professional documents with something to say. But once they reach the hands of people who actually read them and understand the work, it's clear the document is more of a cosplay of a real deliverable someone gets paid for.
"While some people are using AI tools to polish good work, others are using them to create content that is actually unhelpful, incomplete, or missing crucial context about the project at hand."
Based on a survey of 1,150 U.S. employees, 40% reported receiving workslop in the past month.
Colleagues and managers then have to spend extra hours rewriting or fixing it: an "invisible tax" that researchers calculated at $186 per employee per month. In a 10,000-person company, that's more than $9 million a year lost to fake productivity.
Why Workslop Spreads Fast
The mechanism is simple. AI assistants make it cheap to produce something that looks like real work. The cost lands on the receiver, not the sender. The sender saves an hour drafting a memo; the manager spends two hours figuring out the memo doesn't actually answer the question. Multiplied across a company, that's the $186/employee/month figure from the September 2025 HBR study by Niraj Patel, Kate Niederhoffer and the BetterUp research team. The same paper noted that workslop tends to cluster around three task types - status updates, planning documents, and meeting summaries - all of which are exactly where AI assistants are most commonly deployed. The fix isn't banning AI; it's setting standards for what counts as a finished output regardless of how it was produced. Companies that have moved fastest on this (BetterUp's research highlights several) have adopted explicit "AI-aware quality checks" - the document still has to answer the question, even if AI helped write it.
How to Reduce Workslop in Your Office
Define what "done" looks like for the document types AI is now drafting - status updates, briefs, summaries. Anything that doesn't meet the standard goes back, regardless of who or what wrote it.
Train managers to spot workslop signals: vague generalisations, missing specifics, no clear action, perfect grammar with no point.
Default to "AI-assisted is fine; AI-only is not." The senior reviewer must add their judgment before something ships.
Build a culture where sending workslop costs the sender, not the recipient. The cheapest fix is calling it out the first time.
For more on setting boundaries around workplace communication, the same etiquette logic applies to AI-generated content.
