Shopify's Lütke: AI Review Answers 'Didn't Matter' — The Question Did

By James Calloway

Shopify CEO Tobi Lütke says it didn't matter how employees answered the AI question the company added to its performance reviews. What mattered was having it there at all. Speaking on the Sources podcast, he said the question can "give a company a reputation to live up to."

Lütke made the comments to Alex Heath in an episode published Oct. 28, recorded at Shopify's Toronto office. It's his fullest account yet of what happened to the AI policies from his April 2025 memo — the one that declared "reflexive AI usage is now a baseline expectation at Shopify."

What did the April 2025 memo actually require?

Two things. First: "We will add AI usage questions to our performance and peer review questionnaire." Second: teams had to show why AI couldn't do the work before asking for more headcount and resources.

Heath asked whether both rules still stand. Lütke said they now happen on their own. "I don't have to enforce these kind of things," he said.

That's an 18-month arc from mandate to habit — the takeaway being that Shopify sees its AI policies as self-sustaining now, not enforced.

Why doesn't the answer matter?

Here's the odd part. Lütke said he doesn't even know whether the AI question is still in Shopify's current review. He described it as a peer-rating question about how reflexively a colleague shows up with an AI prototype, and admitted "sometimes you play a little bit of tricks there." The answer, he said, didn't matter.

He noted the memo read as controversial outside Shopify but "really not inside of a firm." More internal attention came only after the press "made a big deal out of it" — which he took as proof Shopify had already made the transition.

In short: the question worked as a cultural signal, not a scoring mechanism.

What happens to AI skeptics?

Lütke didn't soften on resistors. He pointed to open-source projects that have banned AI contributions and said he's entertained by debates over whether code was "hand chiseled by humans or somehow computed by math." Good code should be accepted either way, in his view.

His management philosophy: a company should decide who it's for and who it isn't. "Heightening the misalignment with the people you don't want is actually almost more important than having a good careers page," he said.

An AI skeptic who is "just kind of trying to lay low" will never become a top performer where everyone around them uses AI agents, he argued. Companies should make that explicit, he said, because everyone deserves to work somewhere they "either are or have a really good shot at becoming one of the top performers." Once it's clear that's not possible, Shopify wants to help that person find such a place elsewhere.

What about the 'slop grenade' problem?

There's a gap in the review-question logic. In a Knowledge Project episode released Sept. 15, Lütke said AI had made one kind of lazy work worse: Shopify staff call AI output handed to a colleague unread a "slop grenade."

A review question about adoption doesn't tell a team lead whether anyone checks AI drafts before they reach an editor or a client. Adoption and quality control are two different signals, and Shopify is only measuring one.

Earlier in the interview, Lütke said public pessimism about AI won't last, because more people will use it themselves and their stories will wash it away. He didn't say whether the AI question will appear in Shopify's next review cycle — which, given his stance on enforcement, may be the point.