Workslop Cleanup Cost: Hours Lost Fixing Thin AI Drafts

A report that looks finished lands in a shared inbox on a Tuesday morning. The headings are clean, the summary reads with confidence, and the tables line up. Then a colleague starts checking the claims, and the workslop cleanup cost begins to show. The draft turns out to be thin underneath its polish, and the person who received it spends the next stretch of the day verifying facts, asking questions, and redoing sections that should have been solid on arrival. That hidden handoff is the subject of new workplace research, and the findings put a price on a habit many offices have started to treat as normal.

What the survey found

According to research from BetterUp Labs and the Social Media Lab at Stanford, two in five American desk workers surveyed said they had received this kind of polished but hollow AI output from a colleague in the previous month. Each incident took recipients close to two hours to sort out, time spent checking claims, rewriting passages, or starting a task again from scratch. The researchers translated that lost time into money: about one hundred eighty-six dollars a month for each affected employee, and close to nine million dollars a year for a company with ten thousand staff. The workslop cleanup cost, in other words, is not a rounding error. It is a line item that most companies never budgeted, because the cost lands on someone other than the person who sent the draft.

The damage goes past the hours. Recipients in the survey often came away with a lower view of the sender, judging that colleague as less trustworthy and less capable after receiving output that looked ready and was not. Nearly one in five AI users in the survey admitted sending low-effort or low-quality AI content themselves, a confession that points to pressure as much as carelessness. When teams are told to move faster with AI tools and given little guidance on what counts as finished, sending a draft that merely looks complete can feel like the safest way to keep up. The recipient pays the workslop cleanup cost for that choice in time and in confidence.

Why managers are paying attention

A separate global study led by the University of Melbourne with KPMG adds context to the pattern, and it shows why the workslop cleanup cost keeps growing. Two-thirds of respondents in that study said they use AI regularly, and many employees said they had relied on AI outputs without checking whether the output was accurate. Training has lagged behind adoption: fewer than four in ten employees in that study reported receiving workplace training on AI, and only a minority said their workplace had written guidance on generative AI use. Inc. magazine, reporting on the same BetterUp and Stanford survey, noted that the workslop cleanup cost falls hardest on teams with no shared standard for checking AI-assisted work before it moves on.

The fix that researchers and managers point to is unglamorous and specific. Teams need a shared definition of finished work, and a check step before any AI-assisted draft leaves its author. That means verifying names, numbers, and claims against original sources, reading the full draft rather than skimming its summary, and saying plainly when AI helped produce it so the recipient knows what to check. Managers can help by treating that check as part of the job rather than extra work, and by asking for the source material alongside the draft. A short note on what was verified, and what was not, keeps the workslop cleanup cost from landing on the recipient unannounced.

None of this argues against using AI at work. Drafting help, summarising, and brainstorming remain useful, and many workers do produce stronger output with these tools. The research draws a line between using a tool and handing off its raw output as finished work. When the draft is checked before it is sent, the time saving stays with the sender and the team. When it is not, the workslop cleanup cost moves quietly to the next desk, where it is paid in hours, in rework, and in trust that takes longer to rebuild than any single document. Source: BetterUp on AI washing and workslop.