Apple Intelligence has been live in the US since late 2024. China, one of Apple’s most commercially critical markets, has been waiting ever since. That wait is almost over, and the partner making it happen is Alibaba.
According to Engadget, China’s Ministry of Industry and Information Technology has approved the use of Tongyi Qianwen, the family of large language models more widely known as Qwen, along with Tongyi Wanxiang, Alibaba’s AI image and video generation tool. The approvals were first reported by the South China Morning Post, which is itself owned by Alibaba Group, so take the editorial framing with a grain of salt. But the regulatory filings are real, and the timing is not a coincidence.
The updated filings include network access permits specifically for the iPhone Duo and iPhone 18 Pro. The iPhone Duo is set to launch in China next week, and a more capable Siri is one of its marquee features. Without this approval, that pitch would have fallen flat on arrival. Apple had already received clearance to operate Apple Intelligence on the mainland back in July, but it had not actually flipped the switch. The missing piece was getting the underlying AI models approved by regulators, and that is what just happened.
This matters beyond the product launch itself. China has strict and overlapping rules around AI deployment, data localization, and content generation. Foreign AI models from OpenAI or Google, which Apple works with in the US, are not an option in this market. So Apple needed a local partner, and Alibaba was the obvious candidate. Qwen has become one of the more capable open-weight model families globally, with strong benchmark performance and active development across text, vision, and multimodal tasks. It competes with models from Baidu, ByteDance, and Tencent domestically, and has drawn comparisons to Meta’s Llama series internationally.
Reports that Apple was in talks with Chinese AI firms date back to early 2025, with Alibaba mentioned by name at the time. By summer 2025 it was confirmed that Qwen would be the backbone of Apple Intelligence for Chinese users across iPhone, iPad, Mac, and Vision Pro. This week’s regulatory approval is the final formal step before that arrangement goes live.
The broader picture here is Apple walking a tightrope it has been on for years. The company wants to offer a consistent product experience globally, but China’s regulatory environment forces a different architecture. Instead of routing complex queries to OpenAI’s GPT-4o as it does in the US, Apple will route them through Qwen in China. Users may not notice the difference in day-to-day use, but the underlying models, and the companies responsible for them, are entirely different.
For Alibaba, the deal is a meaningful endorsement. Getting Qwen embedded into Apple’s software stack, distributed across millions of iPhones sold in China, is the kind of real-world deployment that builds credibility fast. It also positions Alibaba well as Chinese AI companies increasingly look to demonstrate enterprise and platform-level adoption rather than just releasing benchmark results.
Apple Intelligence has been available in Hong Kong for some time already, but the mainland’s requirements are stricter. Two years is a long time to wait for a feature Apple has treated as central to its hardware pitch. It’s finally arriving, just in time for the next iPhone launch.
