Shopify, Roblox Report AI Costs Outpacing Labor Savings in Q1 2026

Author: Claude, DeepChain TechFlow

DeepChain Summary: The Q1 2026 earnings season for tech giants reveals a new phenomenon: while AI helps companies freeze hiring and cut jobs, its own token consumption and GPU depreciation are eroding gross margins in reverse. Shopify’s subscription gross margin has been suppressed by LLM costs, and roughly one-quarter of Roblox’s full-year profit margin guidance reduction is directly attributed to incremental AI investments. Amazon, Meta, Microsoft, and Google are projected to collectively spend $725 billion on AI capital expenditures in 2026, a 77% year-over-year increase. For the first time, the two sides of AI’s红利—labor savings and compute consumption—are being accounted for on the same earnings report, with the latter clearly dominating.

The first-quarter earnings season is refining the simplistic narrative of "AI replacing labor."

While a group of tech companies have delivered results showing frozen hiring and accelerated product iteration, they are being forced to explain a more challenging issue to investors: rising AI chip depreciation and unpredictable token consumption are eating away at the savings from layoffs.

On May 5, 2026, during the earnings call, Shopify CEO Harley Finkelstein stated that AI now writes over 50% of the company’s code and has enabled Shopify to deliver more than 300 products and features while maintaining a stable employee count. However, management also acknowledged on the same call that gross margins for subscription solutions are being partially offset by large language model (LLM) costs, and this dynamic is expected to continue.

Shopify's Q1 subscription solutions gross margin was 80%, flat compared to the same period last year, but the cost of maintaining this figure is changing.

According to Shopify’s 10-Q filing with the SEC, the cost of subscription solutions increased by 20% year-over-year in the first quarter of 2026 to $148 million, compared to $123 million in the same period last year. Cloud and infrastructure costs—including AI-related usage—saw a standalone increase of $22 million, serving as the primary driver of this cost expansion. Shopify’s Chief Financial Officer, Jeff Hoffmeister, stated on the earnings call that “economies of scale and improvements in support efficiency were partially offset by rising LLM costs, primarily driven by merchant usage of Sidekick, and this dynamic is expected to continue.”

Sidekick is an AI assistant on Shopify’s embedded platform, and its weekly active stores increased by 385% year-over-year this quarter. Merchants created over 12,000 custom apps using Sidekick this quarter, a more than 200% increase quarter-over-quarter, with nearly half of all Shopify Flows generated by AI. AI-driven store traffic grew eightfold year-over-year, and orders from AI-powered searches increased nearly thirteenfold.

However, this surge in usage means an exponential increase in AI inference calls. Each interaction a merchant has with Sidekick and each proactive suggestion generated by the Pulse feature corresponds to a token invoice paid to upstream model providers.

Shopify separates the accounting for "internal AI" and "external AI" for investors: using AI internally to write code and reduce personnel costs is framed as a win in the "cost game," while offering AI products to merchants externally is positioned as a strategic choice to tie infrastructure costs directly to merchant usage. Finkelstein summarized this logic on the earnings call as: "AI is a structural advantage, not just a cost."

Roblox CFO Naveen Chopra explicitly disclosed on the Q1 2026 earnings call on April 30 that approximately one-quarter of the downward revision to the full-year margin guidance stemmed from increased AI investments and adjustments to DevEx (developer payout) for U.S. users aged 18 and above.

Roblox currently runs over 400 AI models on its own and cloud-based GPUs, processing 1.5 million inference requests per second across use cases such as discovery recommendations, communication safety, marketplace recommendations, and 3D generation.

The company’s management is attempting to offset inference costs through business model adjustments. David Baszucki, co-founder and CEO of Roblox, stated on the earnings call that the company’s upcoming project, “Roblox Reality”—a technology capable of running 2K real-time photorealistic video at 60Hz—will not be offered for free. “This will utilize cloud computing resources. We will implement some form of subscription or paid model, so we believe we can offset the costs on the real-time inference side,” Baszucki explained.

Chopra added that the company’s 2026 capital expenditure guidance remains unchanged, with internal data center GPU deployments primarily meeting this year’s inference demands, while some training tasks continue to rely on the cloud. Roblox previously disclosed that by migrating certain AI inference workloads from third-party clouds to its own data centers, it has already achieved a tenfold efficiency improvement in specific workloads such as safety review and content discovery by the end of 2025.

However, Roblox’s full-year guidance for this quarter includes multiple pressures—such as the aforementioned incremental AI investments, fixed cost deleveraging due to lower-than-expected bookings, and the increased DevEx rate to 37.8% for creators of 18+ adult content—ultimately leading the market to reprice its full-year profitability outlook.

Microscopic cases of Shopify and Roblox exist within a larger macrostructural imbalance.

According to data cited by 24/7 Wall St., the combined AI capital expenditures of Amazon, Meta, Microsoft, and Google in 2026 are projected to reach $725 billion, a 77% year-over-year increase. Meta’s annual capital expenditure guidance ranges between $125 billion and $145 billion, implying daily spending of up to $370 million on data center construction; Microsoft’s 2026 calendar year capital expenditure is set at $190 billion, and Amazon has committed $200 billion.

The ratio of this expense to labor costs is extremely imbalanced. Meta’s total labor compensation—including all salaries, benefits, and equity incentives—amounts to approximately $27 billion. Even if Meta were to lay off all its employees tomorrow, the savings would still be less than one-fifth of its 2026 infrastructure spending.

Wedbush Securities analyst Dan Ives estimated in a research report on April 25 that Meta’s upcoming layoff of 8,000 employees will annually free up approximately $2.4 billion in operating expenses, which can only offset about 12% of the incremental depreciation burden in 2026. In other words, nearly ten dollars in labor cost savings are required to fully offset the financial pressure of every dollar spent on AI computing power.

Meta's CFO, Susan Li, framed Meta's workforce reductions during the Q4 2025 earnings call as "building a leaner operating model to help offset our substantial ongoing investments." This statement explicitly positions layoffs as a financial tool for AI capital expenditures, rather than a byproduct of improved productivity.

The biggest beneficiaries of this ledger game are the foundational models and compute suppliers. Microsoft Cloud has maintained a gross margin of 69% despite the expansion pressures on AI infrastructure; OpenAI’s gross margin is externally estimated at around 50%, and Anthropic’s at approximately 60%. Nvidia continues to achieve a gross margin of about 70% in fiscal year 2026.

Meanwhile, application-layer companies—particularly SaaS players that both consume AI and bundle AI capabilities into subscription-based products—are facing a new financial structure: revenue is highly correlated with AI usage intensity, but cost curves are set by upstream model providers, and each model upgrade may introduce additional token consumption.

Tanay Jaipuria noted in his AI gross margin analysis that although the inference cost for individual models is declining at an annual rate of 80%-90%, prices for frontier models remain stable or even rise; application-layer companies that insist on invoking the strongest model for every request are effectively at the mercy of model providers’ pricing schedules for their cost of goods sold (COGS).

Shopify’s approach is to position AI products as a strategic entry point deeply integrated with traffic and merchants, making the growth in inference costs a proxy for platform integration depth; Roblox’s strategy is to separate premium AI experiences from the free tier, forcing users to pay for inference costs. Behind both paths lies the same consensus: relying solely on layoffs to cover AI computing bills is mathematically unsustainable.