After more than two years of regulatory delays, failed partnership negotiations, and a near-collapse with Baidu, Apple is finally releasing Apple Intelligence in China. The announcement came this week in a press release on the company's Chinese press site, confirming that the iOS 27.3 developer beta will arrive November 4, with the public release in December, supporting iPhone 16, iPhone 15 Pro, iPhone Duo, and iPhone Air.
What makes this story more compelling than a simple software launch is the complex geopolitical landscape in which it unfolds. Apple had to navigate a rigorous Chinese regulatory framework that requires AI services to be hosted by local companies, with data kept within the country. The solution was a partnership with Alibaba, whose Qwen model (also known as Tongyi Qianwen) will provide the backend for generative features — an arrangement reminiscent of the OpenAI ChatGPT integration in the rest of the world, but adapted to Chinese restrictions.
What's in the first wave
Apple is being deliberately conservative with the first release. Features include Writing Tools (rewriting, proofreading, and text summaries), Genmoji (custom emoji creation), the photo object removal tool, Live Translation (real-time translation integrated into Messages, FaceTime, and Phone apps), and Visual Intelligence (camera-based analysis of objects and locations). Live translation via AirPods will work with models 5, 4 with active noise cancellation, Pro 2, and Max 2.
But what's notably absent is equally revealing. Siri AI — the crown jewel of Apple Intelligence, with cross-app context and device action execution — won't be available in this version. The Foundation Models features announced at WWDC 2026 are also excluded, which in China would have been based on Google's Gemini. In other words, China gets an Apple Intelligence that the company already considers legacy in other markets — more than a year old.
Why did it take so long?
The story of Apple's entry into China is a case study in the tensions between technological innovation and digital sovereignty. When Apple Intelligence was announced in 2024, the company promised Mandarin Chinese support. The problem was that, to operate legally in China, the AI needed a local provider — and Apple couldn't use its own servers because data needed to stay in China. The initial Baidu deal collapsed in December 2024 when the parties couldn't agree on privacy and data control. Baidu wanted access to training data; Apple didn't.
The switch to Alibaba in 2026 seems to have been the turning point. Alibaba was already an Apple partner in services like Youku and cloud hosting, which facilitated negotiations. Additionally, Qwen is a mature enough model to serve as Apple Intelligence's foundation — the same model already integrated into the Tongyi ecosystem, with more than 100 million users in China. Regulatory approval came in July 2026, with the service registered with China's Cyberspace Administration.
The market impact
China is Apple's second-largest iPhone market, with approximately 900 million users. For Apple, Apple Intelligence represents a way to justify the premium pricing of its devices — an increasingly important factor in a Chinese market where local competition (Huawei, Xiaomi, Oppo) has had native AI for longer. The Huawei Mate 70, for example, has come with the proprietary Honggou AI model integrated into the operating system since 2025.
But there's a deeper dimension here. Qwen's integration into Apple Intelligence means that, for the first time, a Chinese AI model will be directly embedded in a global operating system. This creates an interesting precedent: Apple, an American company, has chosen a Chinese model as the foundation of its AI features for the Chinese market. It's the opposite of technological decoupling — it's a regulation-forced interdependence, yes, but a real interdependence.
What to expect in the coming months
Apple indicated that support for iPadOS, macOS, and visionOS will follow after necessary approvals — meaning China will get Apple Intelligence across all of Apple's devices, but in phases. iPadOS 27.3 is expected to ship with iOS, while macOS and visionOS may take several months longer.
The big question now: when will Siri AI finally reach China? Apple hasn't commented on a timeline, but the Qwen integration with Siri is already functional within Writing Tools — suggesting the model is ready for the full version. What remains is regulatory approval for more advanced features like device action execution and cross-app context.
Meanwhile, Apple Intelligence in China will be, in the company's words, powered by generative models built by Apple — but with Qwen running the backend. A semantic distinction for consumers, but technically significant: Apple is selling the idea of privacy and on-device processing, but it's actually running a third-party model. In the world of technology, China is where fiction meets regulatory reality — and this particular integration is the clearest example of that collision so far.
Sources: Unite.AI, Apple Newsroom China, South China Morning Post
✓ Independent sources cross-checked and verified before publishing