AI-generated 'workslop' is here. It's killing teamwork and causing a multimillion dollar productivity problem, researchers say (www.cnbc.com)

Researchers at BetterUp and Stanford Social Media Lab have coined "workslop" to describe AI-generated workplace output that looks polished but lacks substance — verbose, copy-pasted or incomplete content that forces colleagues to clean it up. Based on a survey of 1,150 U.S. full-time workers, about 40% say they've received workslop in the last month and estimate ~15% of incoming content is low-effort AI output. Recipients spend an average 1 hour 56 minutes fixing or following up on it (roughly $186 per person per month), and researchers estimate a $9 million annual productivity hit for a 10,000-person organization. The phenomenon spans bad code, slide decks, oddly worded emails and reports, and appears especially in professional services and tech as AI use at work has roughly doubled since 2023. Beyond lost hours, workslop erodes trust and team cohesion: many recipients report annoyance, confusion and reduced willingness to collaborate, and most organizations report little measurable ROI from AI adoption. The researchers highlight telltale signs (overly flowery “purple prose,” unnecessary length, missing details) and argue mitigation is organizational: set clear policies, require disclosure when work is AI-assisted, cultivate team norms about task quality, and adopt a “pilot mindset” so tools augment rather than replace human judgment. In short, unchecked generative AI can scale sloppy work rapidly — organizations must pair tools with governance and training to preserve productivity and trust.