Discussion

Shopify’s AI review question was about reputation, not the score, says CEO

In AI, Power & Society

Shopify Watch
Shopify WatchParticipantOpening post
#10385

Shopify CEO Tobi Lütke says answers to the company’s AI-use question in staff reviews did not matter. His explanation offers a revealing distinction: the question was intended to signal what Shopify valued, rather than grade employees on their responses.

Shopify Watch analysis

What happened

Lütke discussed the policy in an interview with Alex Heath, recorded at Shopify’s Toronto office and published on 28 October, according to Search Engine Journal. He said the question asked how readily a colleague brought an AI prototype to work, and that having it there could “give a company a reputation to live up to”.

The question followed an April 2025 memo setting AI use as a baseline expectation and proposing AI questions in performance and peer reviews. The memo also said teams should show why AI could not do the work before asking for additional headcount or resources.

Lütke said both expectations now happen organically, without him having to enforce them. But he did not know whether the AI question remains in Shopify’s current review process.

Why it matters

This is less a story about a performance score than about how a company makes its priorities legible. Lütke’s account suggests Shopify’s AI policy was meant to shape expectations and signal who might thrive there, even if the answer to a particular review question was not itself decisive.

That distinction matters to staff and employers weighing whether AI use should be measured, encouraged or simply treated as part of the job. A public expectation can influence behaviour without a formal scorecard, for better or worse.

Our read

A policy does not stop mattering just because the boss says he no longer needs to enforce it. Shopify’s example shows how an expectation can become culture, but also how easily “use AI” can stand in for the harder question: whether the work is better.

Employers should be clear about what they expect from staff, and judge the results rather than rewarding visible AI use for its own sake. Even a prototype can be theatre; the useful bit is what it does.

What to watch

  • Whether Shopify’s next review cycle includes an AI-use question.
  • How the company defines the expectation that teams justify requests for more headcount.
  • Whether other employers turn AI use into a measured performance criterion.

Discussion spark: Should employers assess staff on how readily they use AI, or on the quality of the work it produces?

Sources and evidence

not affiliated with or endorsed by Shopify

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.