Over the past year, Gina, a communications specialist, has noticed a frustrating trend: colleagues send her AI-generated drafts with a "weird, inhuman tone." Instead of writing themselves, they feed ideas into chatbots and pass the output to her for editing. She now spends hours humanizing this content, often sending it back with requests to "de-AI" it, which sometimes angers coworkers. Gina is drowning in what experts call "workslop"—polished but empty, error-laden AI material that makes people and companies look bad.
The Rise of Workslop
A 2025 US survey found 40% of workers received workslop from colleagues in the past month. A UK report estimated employees spend nearly as much time verifying AI output as using it, with 32% suffering AI burnout. Despite promises that AI would revolutionize work, many find it makes tasks harder. A WalkMe survey showed 66% of respondents said AI doesn't help them work faster or free up time. The Harvard Business Review recently declared workslop is "destroying productivity."
Why Workers Turn to AI
Fred Funck, director of executive coaching at the Centre for Creative Leadership, says exhausted employees under pressure see AI as a "shortcut that feels 'good enough.'" AI removes friction and reduces cognitive load, but the cost often shifts to colleagues. Nick Renner, skills forecasting partner at BPP, notes that "corrective work"—rewriting odd-sounding output—happens informally and is hard to quantify. "When an hour is saved producing a first draft, but 40 minutes is spent verifying and rewriting, the bottleneck just moves from production to review," he explains.
Detecting Workslop
Rosie Wilkins, brand strategist and founder of Brand by Design, sees clients sending AI-generated copy for social media. She spots tell-tale signs: excessive em dashes, overuse of the word "quietly," and short, choppy sentences. "As soon as people notice these, they tell AI to avoid them, so new patterns emerge," she says. Workslop can have real consequences: a Wall Street law firm apologized for AI hallucinations in court documents, and Deloitte refunded part of a fee after an AI-error-ridden report.
The Long-Term Damage
Renner warns that relying on AI erodes core skills like thinking and writing. "Professional judgement is built through struggling with work, not just completing it," he says. Younger workers may never develop these capabilities, while experienced ones risk expertise atrophy. However, the ability to recognize when AI output is subtly wrong will become increasingly valuable. So next time you're tempted to send AI-generated word salad, remember: your colleague knows, and they're likely seething. At least remove those em dashes.
