Apple Intelligence is expected to arrive in China in the coming months following an iOS update. Apple registered its on-device generative AI service with China’s cyberspace regulator last month, according to Reuters, clearing an important regulatory step for the planned launch. Approval of Apple’s proprietary model would make it the first U.S. company permitted to offer its own AI model in the country.
The arrangement represents a change in how Apple is approaching generative AI in China. Rather than relying entirely on domestic providers, Apple has built a model of its own with help from a Chinese technology partner while continuing to incorporate local AI systems. That structure gives the company a greater role in shaping the technology behind Apple Intelligence without removing its dependence on Chinese partners.
Regulation is a central part of that strategy. Generative AI services offered to consumers in China must undergo regulatory scrutiny before deployment, including reviews covering areas such as training data, safety controls and model behavior. Apple’s registration followed a lengthy process of adapting its AI features to local requirements.
The company has previously confirmed its relationship with Alibaba. Alibaba Chairman Joe Tsai said in early 2025 that Apple considered several Chinese companies before choosing Alibaba as a partner. The subsequent work has expanded beyond simply incorporating an existing Alibaba model, with the companies collaborating on the training of Apple’s China-specific LLM.
The partnership also addresses a significant gap in Apple’s product lineup in China. Domestic smartphone makers including Huawei, Xiaomi, Oppo and Vivo have introduced generative AI capabilities on their devices, while Apple has yet to make Apple Intelligence available in the country. Apple’s global implementation also cannot simply be replicated there because ChatGPT, which Apple Intelligence can use for more complex requests in other markets, is unavailable in mainland China.
There are still unanswered questions about how Apple’s model will work alongside Qwen and other systems. Reuters did not establish the model’s size, computing requirements or whether processing will remain entirely on devices or extend to cloud infrastructure. It is also unclear how closely the China implementation will resemble the architecture Apple uses for Apple Intelligence elsewhere.
One indication of how Alibaba technology could appear within Apple’s software surfaced recently when Apple briefly published a Chinese-language guide describing how Mac users in mainland China could connect Qwen with Siri and Writing Tools. The company later removed the document without providing an explanation.
The regulatory milestone does not eliminate the obligations associated with operating generative AI services in China. Approved systems can remain subject to requirements including content reviews, retraining and additional regulatory scrutiny after launch. That means Apple’s China-specific implementation will have to continue operating within rules governing locally available generative AI services.
For Apple, training its own model while retaining Chinese AI partners creates a hybrid approach to bringing Apple Intelligence into the market. Alibaba provides local technical support and Qwen technology, while Apple gains greater control over a core model intended for its products. The remaining question is how those components will be divided once Apple Intelligence becomes available to Chinese users.
This analysis is based on reporting from MacDailyNews.
Image courtesy of AOL.
This article was generated with AI assistance and reviewed for accuracy and quality.
