Workslop looks polished enough to pass along, but it leaves someone else to verify, correct, or redo the work. Hidden rework can erase AI productivity gains, damage trust between colleagues, and cost large organizations millions.
BetterUp Labs and the Stanford Social Media Lab introduced the term in a 2025 Harvard Business Review article. They define workslop as AI-generated work that looks credible on the surface but lacks the substance to meaningfully advance a task.
The problem isn’t AI assistance. It’s what happens when someone treats generation as completion. An employee produces a polished-looking document, summary, presentation, or analysis without checking whether it is accurate, relevant, or useful. The unfinished thinking then moves downstream to whoever receives it.
The work doesn’t disappear; it changes owners.
BetterUp and Stanford surveyed full-time U.S. desk workers. About 40% said they had received workslop in the previous month, and each incident took nearly two hours to resolve. The researchers estimated a monthly productivity cost of $186 per employee. For an organization with 10,000 employees, that adds up to roughly $9 million a year.
Workday’s 2026 research shows the broader rework problem. Employees reported that 37% of the time AI saved them was offset by correcting, rewriting, or clarifying low-quality outputs.
That distinction matters because time to first draft may fall, while total cycle time stays flat or rises. A dashboard that only records the initial speed gain can make a costly workflow look successful.
The researchers found that workslop can change colleagues’ perceptions of each other. About half of recipients thought the sender was less creative, capable, and reliable. 42% considered the sender less trustworthy, and 32% were less likely to want to work with that person again.
That damage can compound across a team. When employees can’t trust colleagues to apply judgment before sharing AI-assisted work, every document becomes a verification task. People spend more time checking assumptions, confirming facts, and deciding whether the person who created the work actually understands it.
A familiar warning sign is a wide gap between executive AI enthusiasm and employee frustration. Leaders see adoption and faster output. Employees see the cleanup. Measuring only the first half of that experience hides the cost.
Workslop is easy to miss when organizations measure activity instead of outcomes. Most AI dashboards track logins, prompts, active users, and time saved. Those metrics show whether employees are using AI, but not whether the work is accurate, useful, or creating rework elsewhere.
A June 2026 Harvard Business Review article described the organization-level consequence as knowledge decay: a decline in the accuracy and quality of organizational knowledge. When weak AI-generated information enters reports, knowledge bases, or operating decisions without adequate review, the problem can spread far beyond the original task.
The key metric is whether the full workflow improved after review, correction, and downstream use.
The answer isn’t to roll back AI adoption. It’s to build the judgment and measurement practices that keep AI assistance from becoming someone else’s cleanup task.
These controls also help leaders distinguish thoughtful AI use from compliance theater. High adoption paired with rising rework isn’t progress. It’s a cost signal.
Larridin helps leaders see the gap between AI activity and effective AI use.
Larridin AI Fluency measures workforce proficiency across tools, teams, and roles. That gives leaders a clearer view of where employees are developing repeatable, effective AI practices and where targeted support may be needed.
Larridin Workflow Intelligence maps how work moves across applications, including sequence, duration, transitions, and friction. This can surface workflows where AI activity coincides with more handoffs, longer completion times, or added review.
Together with an AI measurement framework, these signals help organizations investigate workslop risk without pretending a usage dashboard can judge the quality of every output.
Book a Discovery Call to discuss how Larridin can help measure AI proficiency and downstream workflow performance.
Workslop is AI-generated work that looks polished but lacks the accuracy, context, or judgment needed to be useful. The recipient must verify, correct, or redo it before the task can move forward.
BetterUp and Stanford estimated that each incident takes nearly two hours to resolve and costs about $186 per employee per month. At an organization with 10,000 employees, the annual productivity cost can reach roughly $9 million.
Workslop often develops when organizations encourage broad AI use without defining appropriate tasks, quality standards, or accountability. Employees may mistake a plausible-looking output for completed work and pass the remaining judgment to someone else.
Track total cycle time, rework, corrections, review loops, and employee reports of low-quality AI-assisted work. Compare those signals with AI adoption and proficiency data. Strong usage paired with flat or declining workflow performance is a reason to investigate.
Wondering how much workslop is costing your organization?
Book a Discovery Call to connect AI proficiency, workflow performance, and measurable business outcomes.