4 minBusiness
Shopify CEO Warns AI 'Slop Grenades' Are Creating Extra Work for Colleagues
Shopify CEO Tobias Lütke says employees are using AI to generate unexamined emails and code, creating cleanup work for colleagues. The phenomenon, dubbed 'workslop,' costs the average worker 3.4 hours per month.
Shopify CEO Tobias Lütke is warning that employees are misusing artificial intelligence in ways that create more work for their colleagues, a practice he calls «slop grenades.» In an interview on The Knowledge Project podcast, Lütke said workers are producing unexamined emails and code with AI and failing to take responsibility for the output.
«We call those 'slop grenades' that people toss at each other,» Lütke said. «That's definitely a bad thing.» He explained that AI is meant to help synthesize points in an email, not turn it into «a big missive» that wastes time. «You don't really read it, and now it has to be reviewed by your colleagues, and they are like, 'this doesn't look right',» he said. «You're just letting AI do the work for you.»
The shift in tone marks a notable reversal for Lütke, who last year told employees that using AI was «a baseline expectation» and that they should first prove they «cannot get what they want done using AI» before requesting additional resources. His latest comments suggest that unchecked AI use is undermining productivity rather than enhancing it.
Lütke is not alone in backtracking. Duolingo CEO Luis Von Ahn announced last year that his company would go «AI-first,» evaluating employees on their AI usage and replacing human contractors with AI. But in May, he told Fast Company that he had gotten carried away by AI's strong performance in writing tasks, which ultimately does not match the creativity of Duolingo's people when scaled. «We may need to write 1,000 different stories for people to learn a language, then you'll find that 20% of the things were just pure slop,» Von Ahn said. «Whenever we scale a lot [of] things with AI, we have to really be careful that slop doesn't get through.»
Researchers have coined the term «workslop» to describe the phenomenon Lütke identified: polished-looking AI output that drags down productivity because it requires revision. A survey of 962 American full-time desk workers conducted this year by BetterUp Labs and Stanford's Social Media Lab found that over half of respondents, 52.7%, reported sending workslop to colleagues. The practice was more common in organizations that encouraged AI use.
More than a third of workers, 38%, reported receiving workslop and estimated it cost them 3.4 hours per month on average to revise it. That figure is up from last year's survey, when 40% reported encountering workslop and said they spent two hours reworking it. The 2025 data pegged the cost of revising workslop at $186 per month for single employees and up to $9 million a year in lost productivity for an organization with 10,000 people.
What separates workslop from ordinary low-quality work is that it looks legitimate on the surface while lacking the components that would make it useful. Survey respondents cited well-structured emails with broken links and code that was more complicated than necessary as examples.
The survey also found that relationships suffer when workslop is suspected. Employees said they viewed colleagues who sent it as less competent and less friendly. Among those who had received workslop, 36% reported wanting to avoid working with those colleagues in the future.
The findings suggest that while AI tools can accelerate certain tasks, they also introduce new forms of friction in the workplace. For business leaders who have pushed aggressive AI adoption, the challenge now is ensuring that speed does not come at the expense of quality and accountability.
