Apple has quietly trained its own large language model for the China market with support from Alibaba, according to Reuters reporting confirmed by multiple sources. The proprietary model breaks from Apple's previous strategy of relying entirely on domestic Chinese partners and paves the way for an Apple Intelligence rollout in China in the coming months.
A Dual-Track China AI Strategy
The move creates what Reuters calls a “dual-track strategy”: Apple's own on-device LLM runs alongside Alibaba's Qwen for cloud-side queries, similar to Apple's ChatGPT extension elsewhere. Apple registered its on-device generative AI service with the Cyberspace Administration of China in July, clearing a critical regulatory hurdle. The company also briefly published, then pulled, a support guide explaining how Mac users can connect Qwen AI to Siri and Writing Tools.
First Foreign Company Cleared For A Proprietary China AI Model
According to the report, Apple is the first foreign company approved by Chinese regulators to offer a proprietary AI model in the country. Chinese authorities require every public-facing generative AI service to clear security reviews, an obstacle that has kept OpenAI's ChatGPT and Google's Gemini out of the mainland market. Apple's workaround gives it a scaffold to compete more directly with Huawei and other domestic Android makers that have baked AI into their flagships.
Why It Matters For Global AI Fragmentation
The China model illustrates how geopolitical fragmentation is producing separate technology stacks: global tech giants may increasingly need distinct models, infrastructure partners, and compliance strategies per market. Apple's Chinese revenues, still the company's second largest region, hinge on delivering AI features iPhone buyers now expect. Related coverage: Anthropic's in-house chip push and Turing's Japan-to-US self-driving expansion.
Reporting based on coverage from Reuters, MacRumors and The Verge.