Experts Warn AI Workslop Triggers 3-Point Knowledge Decay in Businesses
Updated
Updated · Computerworld · Jun 23
Experts Warn AI Workslop Triggers 3-Point Knowledge Decay in Businesses
1 articles · Updated · Computerworld · Jun 23
Summary
Harvard Business Review contributors Matthias Holweg and Thomas H. Davenport say unchecked generative AI use is degrading business processes, with “workslop” eroding productivity, quality control, trust and employee critical thinking.
Three risks drive that decay: verification of AI-tainted content, validation of where humans added real value, and entropy as repeated AI iterations drift farther from original ground-truth data.
Hiring shows the problem in practice—candidates can use AI to optimize resumes and even answer interview questions in real time, pushing recruiters toward more structured forms and more on-site screening.
The authors urge companies to restrict AI to tasks with clear value, document why and how it is used, and track source data so outputs can be tied back to authentic, verifiable material.
They argue public LLMs often add little value for enterprise work, while proprietary or smaller models trained on company data can better augment humans and avoid a repeat of the old productivity paradox.
As companies invest billions in AI, why are they facing a productivity paradox and a costly surge in digital 'workslop'?
Is AI creating a 'missing junior loop' by automating entry-level work, threatening to erase the next generation of human experts?
AI Workslop and Knowledge Decay: The Productivity Crisis Undermining 95% of Enterprise AI Investments
Overview
The rapid adoption of artificial intelligence across industries has led to the rise of AI workslop—low-quality, inaccurate, or poorly structured output from AI tools that demands significant human correction or rework. As of June 2026, this issue is increasingly recognized as a critical threat to productivity, as AI workslop can disrupt company processes and create hidden financial burdens. The growing prevalence of AI workslop highlights the need for organizations to address its impact, as the time and resources spent fixing poor AI output often undermine the promised benefits of AI adoption.