Shopify's chief executive has acknowledged that a question built into the company's performance review process — asking colleagues to rate how well peers used AI — didn't actually influence outcomes. He also suggested the question itself created a reputational problem for the company. For a retailer that has spent over a year publicly pushing AI adoption among staff, the admission is notable.
The backstory matters here. In early 2025, Shopify's CEO sent an internal memo stating that employees needed to demonstrate they'd tried using AI before requesting more headcount or resources, and that AI proficiency would factor into performance expectations. That memo circulated widely and was treated as a model for how leadership might formalize AI adoption inside a company, not just encourage it informally.
The peer-rating mechanism was one way that mandate got operationalized: colleagues were asked, as part of standard review cycles, to assess how well others were incorporating AI into their work. According to the CEO's recent comments, those ratings ended up carrying little to no actual weight in evaluations or decisions. He also noted that having the question on the books at all gave outside observers a specific thing to criticize — turning an internal management detail into a public talking point about surveillance culture or performance theater.
Shopify has not announced it's removing the question, and the CEO's comments read more as a clarification of how the system actually functioned than a reversal of the broader AI-adoption push. The company's official stance — that AI fluency is now an expected skill — remains unchanged.
This follows a now-familiar arc in corporate AI messaging. Executives at several companies, including Duolingo and Klarna, have made bold public statements about AI-first operations or AI replacing specific job functions, only to walk back or soften the framing months later once backlash, retention concerns, or practical measurement problems surfaced. The pattern isn't that the AI push disappears — it's that the loudest, most quantifiable version of the mandate tends to quietly lose its teeth while the underlying expectation stays in place informally.
What's genuinely new in this story isn't the AI mandate itself — that's old news by tech-industry standards. What's new is a CEO openly admitting that a formal measurement tool built to enforce that mandate didn't work as intended, and that the attempt to measure it created its own problem.
For small business owners experimenting with similar ideas — adding AI usage to performance reviews, requiring staff to document AI experimentation, or tying raises to tool adoption — this is a useful data point. Measuring something as fuzzy as how well someone used AI, through peer opinion, runs into the same problem measuring creativity or initiative does: it's subjective, hard to standardize, and easy for employees to perceive as either box-checking or surveillance.
If you're considering building AI expectations into reviews, the more defensible approach is tying it to concrete outcomes — tasks completed faster, specific workflows automated, measurable time saved — rather than asking colleagues to rate each other's AI habits. The latter invites exactly the kind of inconsistency Shopify's CEO now says made the metric meaningless in practice.
Watch for whether Shopify modifies or drops the peer-rating question in its next review cycle, and whether other companies that adopted similar AI-usage scoring — reportedly including some in retail and tech — follow with quiet retractions of their own. Also worth tracking: whether any company publishes actual data showing a formal AI-adoption metric changed employee behavior, as opposed to just generating internal friction.
The practical takeaway: a company talking publicly about mandating AI use is not the same as that company having found a reliable way to measure it. Treat those as two separate claims.
