Apple has taken a big step in China by registering its AI service, Apple Intelligence, with the local cyberspace regulator. This registration clears the way for Apple to officially roll out its AI technology in the country.
The Chinese version of Apple Intelligence will be different from what users see elsewhere. It will include AI models developed by Chinese companies Baidu and Alibaba. This move helps Apple follow China’s strict rules on AI and data privacy.
An Alibaba spokesperson explained that their Qwen AI model will be part of the Apple Intelligence setup in China. This model will be built into devices like the iPhone, iPad, Mac, and Vision Pro that support Apple Intelligence in the region.
China’s New AI Safety Standards
China’s Ministry of Industry and Information Technology (MIIT) has started building a safety benchmark for AI. The goal is to make sure AI systems are safe, fair, and trustworthy. This new benchmark will test AI on six key areas: content safety, value alignment, robustness, fairness, privacy protection, and trustworthiness.
The program covers 31 specific safety risks in five main categories. It uses a mix of automated tests and human oversight. This approach aims to control issues like AI hallucinations, data leaks, and attempts to bypass safeguards known as jailbreak attacks.
The National Industrial Information Security Development Research Centre leads this effort. They are inviting companies and experts to join and help develop these safety standards. Applications for participation were due on a recent Tuesday.
Partnerships and Legal Battles in AI
Alibaba is also teaming up with Honor, a company spun off from Huawei in 2020. Together, they plan to deepen collaboration on an AI-powered operating system called Agentic OS. This partnership will be highlighted at the upcoming World Artificial Intelligence Conference (WAIC) in Shanghai.
Honor will host a special WAIC sub-forum titled “From digital screens to embodied intelligence.” Alibaba Cloud posted about creating a new kind of embodied interaction through Honor’s Robot Phone. This device, revealed in March at Mobile World Congress in Barcelona, features a foldable robotic gimbal arm. It has not yet reached mass production or sales.
Meanwhile, Apple is involved in a legal dispute with OpenAI. Apple claimed OpenAI “never responded” to outreach efforts in February. However, reports show OpenAI did reply, but communication ended after a lawyer at Apple confused OpenAI staffers with similar last names.
Apple’s AI Advances in iOS 27
Apple has released an early public beta of iOS 27. This update includes a new Siri AI that can tap into data across a user’s devices. It can answer complex questions and access more information than before.
The new Siri features world knowledge, onscreen awareness, and a Siri mode within the iPhone camera app. Users can also ask Siri to write messages or texts on their behalf. Apple is introducing “Spatial Reframing,” which lets users move a photo after taking it. Another tool can enlarge images by generating the surrounding content.
The iOS 27 beta is still in early testing and available through Apple’s beta program. It marks Apple’s push to improve its AI assistant in a way that feels smarter and more helpful.
All these developments show China is becoming a key battleground for AI innovation and regulation. Apple’s entry, Alibaba and Honor’s partnership, and China’s new safety measures all point to rapid change in the AI landscape there.
Based on
- Apple Intelligence finally gets regulatory approval in China — engadget.com
- China works on AI safety benchmark as regulators target large model risks | South China Morning Post — scmp.com
- Alibaba to team up with Honor in race to build AI agentic devices | South China Morning Post — scmp.com
- Another turn in the Apple v. OpenAI dispute. | The Verge — theverge.com
- iOS 27 beta: Apple releases early version of major new update – so you can make Siri actually useful | The Independent — independent.co.uk
