Updated
Updated · ZDNet · May 20
Zety Says 51% of Professionals Lose Productivity to AI Workslop as 45% Grow More Cautious
Updated
Updated · ZDNet · May 20

Zety Says 51% of Professionals Lose Productivity to AI Workslop as 45% Grow More Cautious

1 articles · Updated · ZDNet · May 20
  • 51% of professionals told Zety that AI “workslop” — polished-looking but inaccurate or weak output — is lowering their productivity rather than improving it.
  • 45% said that experience has made them more cautious about using AI at work, while 57% cited lower trust in AI as the biggest risk and 46% warned of reputational damage.
  • Executives interviewed by ZDNET said companies need two fixes: redefine productivity around AI-first workflows with human judgment layered on top, and measure whether tools save meaningful time.
  • That payoff still requires persistence, they said, because workers who learn to ground AI systems and blend them with human expertise are likely to be more valuable — and choosier about employers’ AI tools.
With most corporate AI investments yielding zero return, is 'workslop' a tech problem or a failure in business strategy?
If AI requires constant human verification, are we creating assistants or just a sophisticated form of micromanagement?

AI Workslop in 2026: How Low-Quality AI Output Threatens Workplace Trust, Productivity, and ROI

Overview

As of 2026, the modern workplace faces a growing challenge called AI Workslop—AI-generated content that looks polished but lacks accuracy and proper human review. Generative AI tools, while designed to boost productivity, often amplify workers’ existing habits of taking shortcuts and doing superficial work. This leads to a flood of low-quality output, which undermines organizational trust, reduces productivity, and can hurt financial performance. The unchecked spread of AI Workslop makes it harder for teams to collaborate effectively and puts extra pressure on employees to review and fix mistakes, highlighting the need for better oversight and responsible AI use.

...