Shopify CEO warns AI ‘slop grenades’ shift work to human colleagues

Art collage, hand with grenade or bomb on green background with copy space. Concept of war and peace. — Image credit: Search Engine Journal

Shopify CEO Tobi Lütke warned that unreviewed AI-generated content, which he termed “slop grenades,” is shifting work and responsibility onto human colleagues within companies.

Lütke made the comments on The Knowledge Project podcast, released Sept. 15, 2026, highlighting the hidden costs and unintended consequences of widespread AI adoption in the workplace.

The CEO defined “slop grenades” as AI output passed to coworkers without proper review, effectively offloading the task of correction and validation. This practice, he indicated, creates an additional burden for human employees.

Shopify itself integrated AI usage into its core operations in April 2025, making it a baseline expectation for employees. The company also ties AI utilization to performance reviews and headcount decisions, according to Lütke.

Up to half of Shopify‘s code changes, known as pull requests, are generated through conversations with an internal AI agent named River, which operates within the company’s chat system, Lütke reported.

The increasing prevalence of unreviewed AI output has led to a significant rise in demand for freelancers specializing in correcting AI-generated content. Freelancer.com reported an 87 percent increase in such demand, while Upwork saw a 70 percent year-over-year rise, according to data from those platforms.

Fiverr experienced an even more dramatic surge, with a 20-fold increase in demand for AI output correction services between 2023 and 2026, Roger Montti reported in the Search Engine Journal.

A September 2025 survey conducted by BetterUp and the Stanford Social Media Lab found that 40 percent of U.S. desk workers had encountered “workslop,” a term referring to poor-quality AI output. Resolving each instance of workslop took nearly two hours on average, the survey indicated.

Lütke emphasized that the issue stems from a shift in responsibility, where the creator of the AI output fails to ensure its quality before presenting it to others. This dynamic, he suggested, undermines productivity rather than enhancing it.

Source: Search Engine Journal