Apple plans roughly $14 billion in capex on AI this year, against a cumulative $900 billion being committed by other tech giants, and the gap is now the strategy rather than the embarrassment. The company unveiled Siri AI on Monday at WWDC 2026, its largest AI launch to date, with new automated capabilities and a partnership with Google Gemini baked into the operating system. The beta ships to consumers later this year. After two years of being called the laggard, Apple is reframing the race on its own terms.
The financial math is the cleanest part of the argument. Apple posted historic iPhone sales last quarter while spending a fraction of what rivals are spending on model training and data center buildouts. It is also collecting App Store fees on the AI apps that competitors distribute through the iPhone, which means a meaningful slice of the $900 billion competitor commitment loops back to Apple as platform revenue.
Craig Federighi, Apple's senior vice president of software engineering, used his WWDC remarks to draw the contrast explicitly, taking a swipe at competitors chasing AI without a clear product thesis.
Key facts
- 01Apple plans roughly $14B in capex this year, against a cumulative $900B being committed by other tech giants.
- 02Siri AI, unveiled Monday at WWDC 2026, embeds Google Gemini for live web answers and ships as a consumer beta later this year.
- 03Apple posted historic iPhone sales last quarter while spending a fraction of what OpenAI, Meta, and Google are pouring into AI.
- 04App Store fees on AI apps mean Apple is earning revenue from the AI boom without owning a frontier model.
The not-so-veiled message is aimed at both Apple's critics and at consumers who are, by several polling measures, increasingly ambivalent about generative AI. Federighi positioned Apple as the company applying AI to a product people already use rather than chasing model benchmarks for their own sake.
He followed it with a mission statement framed around utility rather than capability.
“At Apple, our mission has always been to turn the potential of advanced technology into helpful and intuitive products for everyone.”— Craig Federighi, Apple Senior Vice President of Software Engineering
On the demo floor, Siri AI now surfaces information buried in inbox and message history, reads what is on screen for context, and uses Gemini to pull live information from the web directly to the device. Chat histories persist across Apple devices, matching the table stakes set by ChatGPT and Claude. The features are incremental on their own, but the embedding location is not — Siri sits below the App Store, which is a distribution layer competitors cannot reach.
That is the structural threat to rivals that the spending gap obscures. Every AI app shipped through the App Store reaches users through Apple's payment rails and Apple's review process. Siri reaches users through the home button. For OpenAI, Anthropic, Google, and any consumer AI startup hoping to win the iPhone, an operating-system-level assistant from Apple is a meaningful headwind regardless of which underlying model is doing the inference.
The contrast with competitors is the most instructive part of Monday's announcements. OpenAI ships updates at a relentless pace but oscillates between consumer and enterprise positioning. Meta is committing tens of billions to AI infrastructure without articulating how it ties back to advertising revenue. Both are spending heavily and explaining late. Apple is spending lightly and explaining its restraint as a feature.
The caveats are real. Siri AI is not in consumers' hands yet, and Apple's last attempt at a major Siri overhaul slipped by roughly two years, which we covered last week. The Gemini dependency also means Apple's flagship AI surface relies on a Google model for live web queries, which is an unusual posture for a company that prefers vertical control. If the beta ships rough or the Gemini hand-off feels seamful, the strategy reads as underinvestment rather than discipline.
If Apple can land Siri AI as a competent default assistant on the iPhone, the slow-and-steady posture turns into the most financially sound AI strategy in the industry: minimal model spend, maximal distribution, and a tax on every rival that needs the App Store to reach users. The race Apple is running is not the benchmark race. It is the one where the hardware keeps selling and the platform keeps earning while everyone else explains their capex to shareholders.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
