Apple is changing its approach to artificial intelligence in China. According to Reuters, the company, with the support of Alibaba, has trained its own large language model designed specifically for the Chinese market. This marks a departure from its previous strategy, under which Apple was primarily to rely on models from local partners. Having its own technology could give the company greater control over Apple Intelligence features and how they are integrated into devices.
The move is primarily driven by regulatory considerations. US models, including ChatGPT, are not available in China on the same terms as in the West. In July, the Chinese regulator CAC announced the registration of Apple Intelligence as one of seven generative AI services operating on devices. Whilst this does not grant complete freedom of operation, it removes a significant barrier to the service’s market launch.
The timing is also significant from a business perspective. The Chinese smartphone market contracted by 4.3 per cent year-on-year in the second quarter of 2026. At the same time, Apple’s shipments rose by 24.4 per cent, and the company’s market share increased to 18.1 per cent. Huawei remains the market leader with a 22.6 per cent share. IDC also points to rising component costs and the waning impact of government subsidies, which may further dampen demand.
Apple Intelligence could therefore become a tool for the company to sustain the current recovery and reduce the technological advantage of local competitors. At the same time, a more regional model of AI development is emerging: different technologies and partners in China, and different ones in other markets.
This could result in greater fragmentation of Apple’s global services, but it also reflects a broader trend. Foreign technology firms wishing to operate in China may increasingly localise not only their data and infrastructure, but also their AI models and technology partnerships.
