On June 10, 2024, at its Worldwide Developers Conference, Apple announced Apple Intelligence, described as a “personal intelligence system” woven into iOS 18, iPadOS 18, and macOS Sequoia. Rather than a standalone app or chatbot, Apple Intelligence was positioned as a default feature of the operating system itself: writing tools, notification summaries, image and emoji generation, and a more capable Siri, available across the iPhone, iPad, and Mac to the company’s enormous installed base. CEO Tim Cook framed Apple’s angle as privacy: “Our unique approach combines generative AI with a user’s personal context to deliver truly helpful intelligence. And it can access that information in a completely private and secure way.”
Technically, Apple split the work between models running on-device and a new server tier it called Private Cloud Compute, running on Apple silicon, for requests that need more horsepower without sending personal data to a general-purpose cloud. The pitch was that the most sensitive processing stays on the phone, and that even the cloud fallback is designed so Apple cannot see or retain the data.
The headline partnership was with OpenAI: Apple integrated ChatGPT, powered by GPT-4o, into Siri and system-wide writing tools, so a user could hand off harder requests to ChatGPT without leaving the app they were in or creating an account, with IP addresses obscured and requests not retained by OpenAI. This made Apple Intelligence a notable case of a platform owner shipping a rival’s frontier model as a built-in option rather than building everything in-house.
For business readers, the significance is distribution. Apple does not usually ship first, but when it builds a capability into the OS it reaches hundreds of millions of devices by default. Apple Intelligence signaled that generative AI was no longer a destination users had to seek out; it was becoming an ambient feature of the world’s most widely used consumer hardware. Apple said the features would arrive in beta in fall 2024 in U.S. English on iPhone 15 Pro models and iPads and Macs with M1 or later chips, a reminder that even a default feature ships gradually and on capable hardware.