The Basics: What is AI Workslop?
5 minute read
You’re presenting next year’s business strategy to your team. The deck is sharp and clear with compelling insights and beautiful visuals. AI helped you shape it, and you think it might be your best work yet. You finish presenting and see the team smiling and nodding. There’s a sense that it has landed. Encouraged, you open the floor to questions. But as they begin, something shifts.
Your confidence fades. The polished sentences have disappeared as you realise you’ve presented a strategy you haven’t entirely wrestled with. It’s not that you don’t understand the business, but that parts of this thinking don’t quite belong to you, and you can’t go as deep as the questions now require.
You hesitate – not because the strategy is weak, but because your relationship with it is. Your best presentation yet turns out to be a deck full of AI workslop.
Never Miss A Story
What is AI workslop?
AI workslop is content that looks like good work at first glance, but collapses under scrutiny. It has structure, polish and fluency, yet lacks the substance to meaningfully move a task forward.
It could be an essay that doesn’t make an original claim, or a business report that reads convincingly but offers no real point of view. At its most obvious, AI Workslop is long-winded without depth, circling an idea rather than developing one.
Professor Jeff Hancock from Stanford University describes this style as a kind of “purple prose” – writing that is elaborate and expressive on the surface, but ultimately hollow. What appears thoughtful can often be reduced to a single, unremarkable idea.
As AI tools become more powerful and more accessible, the risk is not just that we produce more content, but that we produce more content that only imitates thinking.
We are being asked to engage with ideas that no one has truly taken ownership of.
Why does it feel so wrong?
Part of the discomfort comes from something deeper than bad writing.
AI workslop sits close to what psychologists call the “uncanny valley”, a phenomenon where something appears almost human, but not quite, creating a subtle sense of unease (think life-like dolls or humanoid robots).
In the same way that a lifelike robot can feel unsettling, AI-generated work can feel off. It carries the rhythms of human thought, but not the weight of it.
We are, often instinctively, able to sense when something hasn’t been fully wrestled with. When the words are there, but the thinking behind them is thin. When something has been assembled, rather than understood.
And so the discomfort is not just aesthetic, it’s relational. We are being asked to engage with ideas that no one has truly taken ownership of.
The human cost of AI workslop
The consequences aren’t just philosophical, they’re practical.
A 2025 Harvard Business Review study found that more than 40% of US-based full-time employees reported receiving AI-generated content that “masquerades as good work but lacks the substance to meaningfully advance a given task”.
In practice, this often creates a quiet but significant burden. Work that appears finished still needs to be rethought, rewritten, or rebuilt entirely. Those who rely less on AI frequently find themselves carrying this invisible labour, correcting and clarifying what should already have been done.
Research from BetterUp Labs and Stanford’s Social Media Lab found that employees spend nearly two hours correcting low-quality AI-generated work, creating an estimated $186 monthly productivity loss per employee.
But beyond time and cost, something else is at stake. When work is produced without ownership, trust begins to erode. Teams become less confident in what they’re reading, and less certain about who is truly thinking.
“There is a growing tendency to slip into “copy-and-paste mode”.”
Kate Niederhoffer
How to fix the slop
AI is not the problem. Used well, it can sharpen thinking, accelerate learning, and open up new creative possibilities. The risk emerges when it replaces the very work it is meant to support.
As Kate Niederhoffer notes, there is a growing tendency to slip into a “copy-and-paste mode”, allowing the tool to do the work, rather than using it to augment our own capabilities.
The alternative is not to reject AI, but to change our relationship with it.
To treat it as a collaborator, not a substitute.
To use it to challenge our thinking, not bypass it.
To remain transparent about where it has shaped the work, and where we have.
As Brenna Spain writes in The Beautiful Truth, “Skills like critical thinking, creativity, emotional intelligence, and adaptability are no longer ‘soft’ – they are essential.”
The challenge ahead is not simply technological, but human. Because in the end, the real risk of AI workslop is not that it produces bad work. It’s that, over time, it distances us from the act of thinking itself, leaving us with output that looks like insight, but no longer belongs to us.
