Apple May Use Google Cloud for Upgraded Siri

This isn't the first time Apple has leaned on its rival's cloud infrastructure. Apple has been a major Google Cloud customer for years, reportedly spending around $300 million in 2021 for services including iCloud storage. That year, Apple held over 8 exabytes of data on Google's servers, making it the cloud provider's largest corporate client. The potential deeper partnership for Siri follows Apple's decision to use Google's Gemini AI models for a significant overhaul of the voice assistant. This move is part of a broader effort to make Siri more conversational and capable of handling complex tasks, an area where it has lagged behind competitors. While embracing a cloud partnership, Apple is also making massive investments in its own AI infrastructure. The company has embarked on a four-year, $500 billion plan that includes building a 250,000-square-foot server manufacturing facility in Houston, Texas, and expanding its data centers. This dual approach highlights a core tension in Apple's AI strategy. The company has long championed on-device processing for privacy and speed, a method powered by its custom-designed Neural Engine in its chips. However, the immense computing power required for cutting-edge generative AI has pushed the famously self-reliant company toward a hybrid model. Internally, Apple has been working on its own server chips under the codename "Project ACDC" (Apple Chips in Data Center). The goal is to create specialized silicon for AI processing within its own data centers, extending its successful custom chip strategy from iPhones and Macs to its server farms. Despite these long-term investments, Apple's capital expenditure on AI-related initiatives in fiscal 2025 was about $12.72 billion, significantly less than competitors like Google, Microsoft, and Amazon, who spend between $30 billion and over $100 billion. This more measured spending underscores its reliance on partners like Google to bridge the immediate gap in the AI arms race.