Harvard Study: AI-Generated ‘Workslop’ Costs Companies £186 Per Employee Monthly

Researchers from Harvard Business Review have revealed that poorly implemented AI-generated content, termed “workslop,” is costing organisations an average of £186 per employee each month whilst damaging workplace relationships and productivity. The comprehensive study challenges the assumption that increased AI adoption automatically translates to improved business outcomes.

Context and Background

The research, conducted by Kate Niederhoffer, Gabriella Rosen Kellerman, and colleagues, introduces the concept of “workslop” – AI-generated content that appears professional but lacks the substance to meaningfully advance tasks. This phenomenon has emerged as AI usage at work has doubled since 2023, yet 95% of organisations report no measurable return on their AI investments.

The study found that 40% of employees have received workslop content within the past month, with AI-generated material comprising an average of 15.4% of workplace content. Each instance of workslop requires approximately 1.56 hours of additional productive time from recipients to process, evaluate, and respond appropriately.

Beyond the quantifiable time costs, workslop significantly impacts professional relationships. Half of recipients view creators of AI-generated content as less creative, capable, and reliable, whilst 42% consider them less trustworthy and 32% are less likely to collaborate with them in future.

Looking Forward

The findings highlight the critical need for strategic AI implementation rather than indiscriminate adoption. Researchers recommend that leaders develop clear guidelines and norms around AI usage, promoting a “pilot mindset” that emphasises thoughtful experimentation over wholesale deployment.

The study suggests reframing AI as a collaborative tool rather than a productivity shortcut, encouraging organisations to focus on purposeful implementation that creates genuine value rather than simply increasing output volume. This approach could help realise AI’s potential whilst avoiding the hidden costs associated with poorly executed automation.

Source Attribution:

Share this article