Apple spent the first years of the generative AI boom looking late. Microsoft tied itself to OpenAI, Google rebuilt its products around Gemini, and Amazon and Meta committed enormous sums to chips and data centers. Apple introduced Apple Intelligence slowly, delayed major Siri features and relied on outside models where its own software fell short. The company appeared to be losing the most important technology race in years. Its stock now sits near a record high and has become the best-performing member of the Magnificent Seven this year, according to Axios. Investors are now asking whether the largest AI budget also produces the strongest business.
Amazon, Microsoft, Google and Meta are spending at a scale that was difficult to imagine before ChatGPT. The money goes into Nvidia processors, networking equipment, land, power contracts and data centers that can take years to build. Reuters has reported that AI infrastructure is now one of the main forces driving capital spending across the largest US companies. The hyperscalers are moving cash that once supported margins and shareholder returns into equipment that depreciates quickly and must be replaced as faster chips arrive. Their cloud businesses give them a way to sell that capacity, yet the costs arrive years before the returns become clear.
Apple has avoided much of that bill because its AI strategy serves a narrower purpose. It trains its own models, runs Private Cloud Compute and designs chips capable of handling AI workloads. Smaller tasks can run on the device, private requests can move to Apple’s servers, and outside models can cover capabilities Apple has yet to build well. Apple only needs AI to improve the devices and services it already sells. It has little reason to build a global cloud business on the scale of Microsoft, Amazon or Google.
That approach looked weak when investors focused on model rankings, chatbot launches and benchmark results. The picture changes as infrastructure spending begins to pressure cash flow. Hyperscalers are sending a growing share of their money to chipmakers, utilities and construction companies, while their own shares have often received a smaller reward. Reuters reported this week that confidence in AI remains high even as concern over valuations and spending continues to rise. Investors still believe AI will create enormous value. The open question is how much of that value will remain with the companies paying for the infrastructure.
Apple already owns the users
Most AI companies must build a model and then find distribution. Apple starts with roughly 2.5 billion active devices and direct control over the operating systems people use throughout the day. It can place AI inside Messages, Mail, Photos, Siri and third-party apps without asking users to adopt a separate platform. It also controls the hardware, permissions and personal context that make an assistant useful. An outside model may generate the answer, while Apple still controls where that model appears and what information it can access.
This gives Apple a different path to revenue. The company does not need to charge for every request or sell cloud capacity to justify its AI spending. AI can help sell more expensive iPhones, improve retention and strengthen services. Microsoft, Amazon and Google depend more heavily on direct returns from cloud infrastructure. Apple can spread the value across products it already knows how to monetize.
Apple still has serious weaknesses. The first Apple Intelligence rollout was uneven, and delayed Siri features damaged the company’s credibility. Dependence on external models also creates long-term risk. OpenAI, Google and other AI companies are building agents, browsers and hardware that could compete for the relationship Apple currently owns. A persistent AI assistant could eventually become more important than the operating system beneath it, weakening Apple’s control over distribution.
That threat remains distant. The iPhone is still the main device through which most users will experience consumer AI. Apple can improve its position without matching the infrastructure spending of its rivals. It has time, strong margins and the option to buy or partner for capabilities it cannot build quickly enough.
Apple has yet to prove that its AI strategy will deliver the best products. Its financial position is already becoming clearer.
The company looked weak when the market measured AI by model launches. It looks stronger when the market measures who is paying for them.
