Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

MacRumors

macrumors bot
Original poster


Apple briefly published a support guide in China explaining how Mac users can connect Alibaba's Qwen AI to Siri and Writing Tools, before pulling it entirely less than a day later.

apple-intelligence-black.jpeg

Reuters first spotted the Chinese-language guide on Apple's support site on August 8, which explained how eligible Mac users in mainland China could connect Qwen, Alibaba's family of generative AI models, to Siri and Writing Tools. According to the document, the extension required macOS 26.6 or later, needed users to sign into their own Qwen account, and specified that Alibaba was barred from using submitted materials to train or improve its models.

Siri would be able to lean on Qwen for more detailed responses to certain requests, including analysis of photos and documents, while Writing Tools could use it to generate text and images. No equivalent guide was published for the iPhone or iPad.

The setup closely mirrors Apple's existing ChatGPT extension available elsewhere, which similarly lets Siri hand off complex requests and gives Writing Tools a "Compose" option, all on an opt-in basis.

By August 9, the guide had been removed. Chinese news aggregator ReadHub reported that the guide had disappeared from Apple's site, and TechNode reports that Apple's customer service told Chinese media it had received no notification of a new feature launch. Separately, tech analyst Liu Dingding told the Global Times that the takedown could reflect pending technical adjustments still tied to the regulatory process, rather than a full reversal.

It's a similar pattern to what happened back in March, when Apple Intelligence itself briefly went live for some Chinese users before being pulled offline without regulatory approval. China maintains strict rules around releasing AI features, and in practice requires companies like Apple to partner with local providers. Apple originally lined up Baidu as its primary China AI partner, but by February 2025, reports indicated Alibaba had taken over as the lead partner after Baidu's models fell short of Apple's requirements, with Baidu shifted to a supporting role.

Chinese regulators cleared Apple Intelligence for launch in the country in July, registering Apple's on-device AI service with the Cyberspace Administration of China. That approval covered the iPhone only, and Apple did not say whether or when the iPad and Mac would follow, making this week's brief Mac-specific guide the clearest signal yet on the Mac timeline specifically.

It remains unclear when Apple will formally relaunch the Qwen integration for the Mac, or whether the guide's brief appearance points to an imminent rollout.

Article Link: Apple Briefly Posts China Guide for Connecting Siri to Qwen AI
 
  • Angry
Reactions: Z-4195
I wonder which PR speak covers this situation:
1. From time to time, Apple removes documents. Apple generally doesn’t discuss their internal deliberations.
2. At the very last minute, Apple discovered X, so Apple had to pull the guide.
 
How about allowing US users to connect to our own local LLM running on Apple Silicon? 🙄

nobody is stopping you from installing Ollama, LM Studio or llamacpp.

What is very difficult is making Siri work reliably with thousands of different models which each have their own issues and bugs. Some of them are even trained to behave a little naughty.

A corporation’s support department cannot handle all that.
 
  • Like
Reactions: JosephAW
nobody is stopping you from installing Ollama, LM Studio or llamacpp.

What is very difficult is making Siri work reliably with thousands of different models which each have their own issues and bugs. Some of them are even trained to behave a little naughty.

A corporation’s support department cannot handle all that.
I’ve always thought that, as the ONLY company supporting their products, some of their decisions come down to “How many people would love this enough to drop everything and switch (or buy ours first) versus how many more support calls are we going to get.” People think multiple volumes or multiple models are a nice to have, but it wouldn’t bring in enough new users to be worth the price of support!
 
nobody is stopping you from installing Ollama, LM Studio or llamacpp.

What is very difficult is making Siri work reliably with thousands of different models which each have their own issues and bugs. Some of them are even trained to behave a little naughty.

A corporation’s support department cannot handle all that.
Its not difficult. It is called Siri Extensions and Apple is already working on it, what's not clear is if the extensions are allowed for local model servers.

If Siri can work with ChatGPT, Claude, Gemini, and now Qwen, it can work with just about any model from the past 2-3 years, because they all speak the same language. An OpenAI-compatible /v1/chat/completions endpoint. Ollama, LM Studio, llama.cpp server mode, vLLM, they all expose this.

Apple already normalizes across four different vendors with different auth and quirks. Pointing that same layer at localhost:11434 instead of a cloud endpoint isn't a different problem, it's the same client code hitting a local port.
 
  • Like
Reactions: expensiveram
I wouldn't be surprised if Qwen AI is directly tied to the CCP.
It's based in China. It makes no difference if it's fully owned and operated by the CCP or fully owned and operated by a private citizen; the effect is the same. The CCP has the authority to obtain any information they want from their domestic corporations.

This is not speculation; it's been widely reported. It's also the primary reason that the US government takes issue with US companies outsourcing to China.
 
  • Like
Reactions: inventor7777
It's based in China. It makes no difference if it's fully owned and operated by the CCP or fully owned and operated by a private citizen; the effect is the same. The CCP has the authority to obtain any information they want from their domestic corporations.

This is not speculation; it's been widely reported. It's also the primary reason that the US government takes issue with US companies outsourcing to China.
But open weight models like Qwen can be downloaded and run offline, how does an offline model send data to China.
 
  • Like
Reactions: I_love_money
But open weight models like Qwen can be downloaded and run offline, how does an offline model send data to China.
It doesn’t, that’s the advantage of running local models. By the way, not only the Chinese DeepSeek, but ChatGPT and Claude also save all the information you share with their models. That’s why you should never share sensitive information with any not fully local model.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.