Apple

Apple Inc. is the world's most valuable company by market capitalization, known for its integrated hardware-software ecosystem spanning iPhone, iPad, Mac, Apple Watch, Vision Pro, and associated services. It was founded in 1976 by Steve Jobs, Steve Wozniak, and Ronald Wayne.

Apple's AI strategy centers on Apple Intelligence, a personal intelligence system deeply integrated across iOS, iPadOS, and macOS. Apple Intelligence includes on-device language models for writing tools, image generation (Image Playground and Genmoji), and a substantially enhanced Siri with conversational context, on-screen awareness, and the ability to take actions across apps. For complex queries that exceed on-device capabilities, the operating system routes selected requests to OpenAI's ChatGPT — a partnership that has been the subject of ongoing contractual dispute between the two companies. At WWDC on June 8, 2026, Apple confirmed that the next generation of its Apple Foundation Models — and the rebuilt, chatbot-style Siri they power — run on a custom Google Gemini model (reported at roughly 1.2 trillion parameters under a multi-year deal worth about $1 billion a year) hosted inside Private Cloud Compute, with Apple's contract barring Google from training on Apple users' queries. A new Extensions system lets users route Siri to Claude, Gemini, or Grok, ending ChatGPT's former exclusivity as the only external option. The rebuilt assistant, branded Siri AI, reached the public on September 14, 2026 with iOS 27 as an opt-in beta, initially in English and not available on iPhone, iPad or Apple Watch in the European Union. Apple's release notes attach daily usage limits to features that run on its servers, Siri AI among them, and say expanded access will be offered for a fee in the future, which would make it the first paid tier of Apple Intelligence. Apple has not said what it will cost.

Apple differentiates through its privacy-first approach to AI: most Apple Intelligence features run on-device using the company's custom Neural Engine silicon, with Private Cloud Compute handling tasks that require more processing power while maintaining cryptographic privacy guarantees. Apple's M-series and A-series chips, with dedicated Neural Engines delivering up to 38 TOPS, make every Apple device an AI inference machine — a unique advantage over cloud-dependent competitors.

Apple's AI strategy has also produced the first major legal precedent for AI marketing claims: a $250 million class-action settlement covering eligible US buyers of iPhone 15 and iPhone 16 (purchased between June 10, 2024 and March 29, 2025) over allegations that Apple misrepresented the availability of Apple Intelligence — particularly an upgraded Siri marketed as ChatGPT-class — at launch. Eligible buyers can claim up to $95 per device; Apple admitted no wrongdoing. The settlement and the company's broader multi-vendor pivot for Siri — existing ChatGPT routing in the OS, new Gemini-powered standalone app, Apple Intelligence on-device — mark Apple's transition from sole-AI-partnership integrator to a vendor-agnostic AI distributor, with iPhone, iPad, and Mac as the world's most consequential AI distribution surface.

Apple's other AI position is hardware rather than models, and it is the one place the company sells directly to AI developers. On August 25, 2026 Apple announced the M6 and M5 Ultra chips alongside a rebuilt Mac Studio and Mac mini, and said explicitly that the top configuration runs large language models with hundreds of billions of parameters entirely on device. The M5 Ultra pairs up to 512 gigabytes of unified memory with 1.2 terabytes per second of bandwidth — double the 256 gigabytes of the M3 Ultra it replaces, since Apple skipped the Ultra tier in the M4 generation — and Apple claims 4.5 times the peak graphics compute for AI against that M3 Ultra. Mac Studio starts at $2,499 with the M5 Max and $5,499 with the M5 Ultra. The strategic point is capacity rather than raw speed: unified memory lets a single desktop hold models that no consumer discrete-GPU workstation at the same price can load at all, which has made the Mac the default machine for private, local inference work.

In Greater China, Apple Intelligence runs on a different stack. China's Cyberspace Administration cleared the feature set in July 2026, following a filing that named Alibaba's Qwen as the model handling text and image understanding and generation across iOS, iPadOS, macOS, and visionOS for Chinese users, with Baidu contributing a smaller model alongside it. Apple had explored partnerships with Baidu, DeepSeek, and ByteDance before settling on Alibaba, and had spent roughly two years unable to ship in the market where it earns its largest revenue outside the United States. Regulatory clearance is a precondition rather than a launch date — the rollout still depends on security review, engineering adaptation, and operating-system updates.