John Ternus delivered record revenue in his first earnings report as Apple's chief executive. Apple reported fourth-quarter revenue of $114.5 billion, up 12% year over year, beating the $113.4 billion analyst consensus2,8. Earnings per share came in at $2.10 against a $1.99 forecast. And the market punished him for it. Apple's AI buildout is compressing the margins that underwrite its valuation, and Ternus's first forward guidance made that compression explicit. The collision between Apple's AI investment timeline and its margin expectations is the defining tension of the Ternus era, and the Q1 gross-margin guide is the first hard number attached to it.
iPhone revenue hit a September-quarter record of $57.6 billion, up 17%, and Mac revenue surged 24% to $10.8 billion. Full-year revenue reached a record $479 billion12. None of it mattered once CFO Kevan Parekh delivered the December-quarter gross-margin forecast: 45% to 46%, trailing the 46.3% analyst estimate7,4. Parekh attributed the squeeze to two factors: higher memory costs and costs related to the ramp of Siri AI infrastructure. The distinction matters. Memory is a cyclical commodity input that will eventually ease; Siri AI infrastructure is a discretionary capital commitment that scales with usage. Apple listed them as co-equal drivers, and only the second is under management's control.
Ternus himself underscored the severity. "What we're seeing with memory pricing right now is unprecedented," he said on the call, predicting that memory costs in the December quarter would be even higher than in the prior three months9. The stock, which had closed at $338.32, fell as much as 4.4% during the earnings call before ending after-hours trading at $328.81. By the next morning, AAPL was down more than 4% in pre-market trading3.
ANALYSIS The margin guide is notable not because memory prices are a surprise to anyone following the industry, but because it quantifies how directly Apple's AI ambitions feed through to the income statement. The company acknowledged that Siri AI data-center costs are part of the margin drag alongside memory. Apple is now absorbing both a cyclical commodity squeeze and a structural infrastructure buildout at the same time, and Ternus's guidance suggests the company has limited near-term ability to offset either cost.
Apple did raise prices on iPads and MacBooks in June and on iPhones in September, and Ternus said those increases "provided some offset". The September-quarter gross margin came in at 47.9%, but roughly one percentage point of that reflected tariff refunds, most of which have now been paid10. With the tariff benefit largely behind it and memory costs still climbing, the sequential step-down to a 45%–46% guide represents a visible tightening that management tied explicitly to AI-related spending, not to a one-time event.
The earnings call itself drew pointed criticism. Bloomberg's Dave Lee described it as "a timid affair, with Ternus taking cover behind a carefully written script as analysts peppered him with questions"6. Lee argued that Ternus missed a chance to articulate a vision for Apple under new leadership, what Lee called "the fundamentals of the Ternus Doctrine". Ben Barringer, head of technology research at Quilter Cheviot, called the results "OK, if on the disappointing side".
The criticism of Ternus's communication style, while subjective, intersects with a substantive problem: investors are being asked to absorb margin compression without a clear narrative about what the spending buys. The two most visible AI-linked initiatives, Siri AI infrastructure and the robotic HomeView line, are both cost centers with uncertain payoff timelines. The first hits cost of goods now; the second hits R&D and is almost certainly immaterial to the current quarter's margin guide. They belong to different time horizons but the same strategic question: what is the return profile of Apple's AI spending?
On the hardware side, The Information reported that Apple is developing a robotic home device, code-named J595, with a 9-inch screen on a swiveling arm, targeting a price of around $1,0005,1. The device could be branded HomeView Ultra, positioning it above the $299 HomeView that ships this month13,15. Apple is aiming for a 2027 launch, though development delays may push it to 2028. The project dates back to 2018 in its earliest prototype form and received approval from former CEO Tim Cook to continue. Aaron Tilley at The Information reported that the device was originally envisioned mostly for video calls, but ambitions expanded once Apple executives began wanting to add AI capabilities14.
The base HomeView is priced at $299. The robotic version's reported $1,000 price would be more than three times that entry point. The device illustrates the pattern on a longer time horizon: Apple's AI aspirations are inflating both the cost and the timeline of products that were conceived with more modest goals. Its R&D costs are not what is compressing this quarter's gross margin, but the project is the clearest example of how AI scope creep extends Apple's investment-to-revenue cycle.
Ternus, who took over from Cook on September 1, inherits a company whose operational excellence is not in question. Revenue guidance for the December quarter of 7% to 10% year-over-year growth topped the 6% analyst estimate. Ternus called the iPhone Duo "the most transformational iPhone since the original"11. The top line is growing. The question is whether the bottom line can keep pace as AI costs scale.
The revenue beat and the margin miss together frame the central question of the Ternus era. Apple can still grow the top line through pricing power and new form factors. But its AI investments, from server-side Siri infrastructure to a robotic home device that has been in development since 2018, are now visibly eating into the margins that sustain a valuation built on operational efficiency. Ternus's first quarter showed he can hit the revenue targets Cook set. His next quarters will show whether he can articulate a return on the AI spending that justifies the margin cost.
Apple's December-quarter gross margin will be the first marker: whether the 45%–46% guide holds or worsens when results land in late January 2027. Any update on the robotic HomeView device's timing at Apple's spring event cycle would indicate whether the 2027 target is real or slipping. And if Apple begins disclosing Siri AI infrastructure spending as a separate line or category, that would mark a shift toward the transparency investors did not get on this call.