Workslop: When AI Hurts Productivity Instead of Helping
Harvard Business Review research shows how AI-generated ‘workslop’ clogs workflows and destroys collaboration. Here’s why careless adoption is costing companies millions — and how to avoid it.

AI isn’t just about efficiency and automation. Used carelessly, it can generate what Harvard Business Review researchers call “workslop” — output that looks polished but lacks substance, shifting the burden onto colleagues.
The September 2025 HBR study (in collaboration with Stanford and BetterUp Labs) reveals the scope of the problem: 40% of employees report receiving workslop monthly. Each instance wastes almost two hours. For an organization with 10,000 employees, that invisible tax amounts to more than $9 million a year in lost productivity.
Workslop isn’t just an annoyance — it corrodes collaboration. Employees describe frustration, confusion, and even mistrust when asked to rework AI-generated reports, slides, or emails. Colleagues who send workslop are often perceived as less capable, less creative, and less reliable. Over time, this damages trust and teamwork, undermining the very adoption efforts companies hoped AI would accelerate.
Workslop Makes Colleagues Think Less of Each Other.
Here’s the simple truth: If you just use AI without using your brain, the result isn’t productivity — it’s workslop.
The researchers point to three drivers:
- Indiscriminate mandates that tell employees to “use AI everywhere” without guidance.
- Passenger mindsets, where workers lean on AI to dodge effort instead of enhance creativity.
- Weak collaboration norms, where AI outputs are dumped into workflows without context or quality control.
The solution? Leaders must model purposeful AI use, set clear guardrails, and emphasize collaboration. AI should be framed as a partner, not a shortcut. Teams that adopt a pilot mindset — high optimism, high agency — use AI to accelerate real outcomes, not generate busywork.
At ReplyFabric, we take this lesson to heart. Our AI agent isn’t designed to create more noise. Instead, it tackles routine, repeatable tasks in shared inboxes — categorizing, drafting, and routing emails — with dual validation (AI in the loop + Human in the loop) and context awareness. No workslop, no invisible tax. Just measurable productivity gains and the freedom for humans to focus on high-value work.
Because AI should never dilute human work. It should elevate it. That’s the Human Premium.
👉 Read the full Harvard Business Review article (paywall)
👉 Read an artical about the article on CNN
Frequently Asked Questions

About the Author
Tom Vanderbauwhede is the founder & CEO of ReplyFabric, lecturer in AI at KdG University, and a seasoned entrepreneur with 25+ years of business experience. He holds master's degrees in Applied Economics, Business Administration (MBA), and Strategic Change Management & Leadership. Tom is passionate about building AI tools that reduce email overload and help teams focus on what matters.
Connect with Tom on LinkedIn and follow his journey as a founder.