Key facts
- Apple is reportedly developing a proprietary AI model for China in partnership with Alibaba.
- This strategy allows Apple to introduce AI features in China despite restrictions on foreign models.
- The move is seen as vital for Apple's market position and US soft power in China.
- The debate between open-weight and closed AI models has historical parallels, with potential for coexistence.
- Major tech firms advocate for a broad, open AI ecosystem to foster competition and distribute benefits.
Apple is reportedly developing its own proprietary AI model for the Chinese market, collaborating with Alibaba to navigate the country's protectionist technology landscape. This move comes as US AI models like ChatGPT and Claude are not readily available in China, and it follows Apple's work with Google on its global "Apple Intelligence" features.
The partnership between Apple and Alibaba for AI features was submitted for approval to China's Cyberspace Administration in February 2025, and the regulator has since registered the service. Apple may also integrate technology from Baidu. This strategy is considered strategically vital for Apple, helping to maintain its popularity and US soft power in China, especially as Chinese consumers show significant interest in generative AI, with its absence impacting purchasing decisions.
This development occurs amidst a broader debate in the tech industry regarding open-weight versus closed AI models. Analysts like Deutsche Bank's Adrian Cox have drawn historical parallels to technological format wars, suggesting that a coexistence of both open and closed models is likely, similar to the smartphone market's iOS and Android ecosystems. Critics of open models raise concerns about safety and potential misuse, while proponents emphasize the benefits of competition, cost reduction, and broader distribution of AI's advantages.
Major technology companies, including Nvidia, Google, Meta, Microsoft, and OpenAI, have jointly advocated for the development of a robust, open AI ecosystem. They argue that open-source software historically fostered innovation and institutional sovereignty, and that maintaining access to open models strengthens defenses and competition, preventing AI benefits from being concentrated in a few hands. Nvidia CEO Jensen Huang has also voiced support for the necessity of both open and closed frontier AI models.